A flutterometer is being developed for use as a data-analysis tool for flight flutter testing. The tool is based on an on-line implementation of a robust flutter-analysis technique, called the "* method," that was described in "Characterizing Worst-Case Flutter Margins From Flight Data" (DRC-97-03), NASA Tech Briefs, Vol. 21, No. 4 (April 1997), page 62.

Flight testing is performed to evaluate the characteristics of an aircraft and to determine a range of flight conditions, called a "flight envelope," in which an aircraft can be operated with safety. A particularly dangerous instability that one must avoid is flutter, which results from coupling between structural and aerodynamic forces. Flight flutter testing is flight testing for determining a safe flight envelope in which an aircraft will not exhibit a flutter instability.

Figure 1. The NASA F/A-18 Systems Research Aircraft, here shown high above Edwards Air Force Base, was mathematically modeled in a computational simulation to demonstrate the performance of the flutterometer.

In traditional methods of flight testing, one expands a flight envelope via a series of test points, using measurements (that is, sensor outputs) generated in response to some excitation. Such dynamical properties as damping are estimated from the measurement data, and trends are used to determine whether the envelope may be expanded further. These methods are inefficient, and to be able to use these methods, one must ensure that data points are closely spaced, inasmuch as the trends do not guarantee which increases in flight conditions may be safely considered. Also, flight flutter testing is potentially dangerous to the crew because an unexpected flutter instability (one not predicted from the trends) can occur during a transition to a new test point.

The * method provides for computation of worst-case flight conditions for which flutter may occur, by consideration of errors and uncertainty in a mathematical model. The flutterometer uses flight data to identify errors and uncertainty in a model, so the resulting worst-case flutter margins predicted by use of the * method directly account for data that describe the aircraft (as distinguished from a mathematical model of the aircraft).

Figure 2. These Plots Show the Dynamic Pressures at which flutter was predicted to occur, as a functions of time, according to the flutterometer and to two mathematical models of the F/A-18 airplane.

The flutterometer can reduce the risk and cost associated with flight flutter testing by predicting unstable flight conditions at a current test point in order to determine what increases in flight conditions may be safely considered for the next test point. The envelope expansion can proceed more rapidly, inasmuch as test points can be chosen on the basis of traditional, but unreliable, damping trends along with the additional information from flutterometer predictions.

A flight flutter test of the F/A-18 Systems Research Aircraft (see Figure 1), in which the envelope at mach 1.2 was expanded by decreasing altitude and increasing dynamic pressure, was simulated computationally to demonstrate the performance of the flutterometer. The F/A-18 mathematical model used in the simulation provided for a representation of temporal variation of mass corresponding to consumption of fuel, and for representation of measurements by wing-tip sensors. Another, nominal mathematical model of the F/A-18 airplane that includes 10-percent errors in structural stiffnesses and was formulated for constant mass corresponding to full fuel was used to predict unstable flight conditions for the simulated F/A-18.

Figure 2 shows the dynamic pressure at which flutter occurred as the simulation progressed. The dashed line represents the unstable flight conditions for the F/A-18 simulation; this line shows changes, even at constant-mach-number flight, because of the temporal variation of mass. The dotted line represents the traditional prediction, which was inaccurate inasmuch as it was based on the nominal model with no accounting for errors. The solid line represents the flutterometer prediction, which was computed by use of sensor data starting at 3 minutes (simulated time) after the beginning of the simulation to identify errors in the nominal model. These errors account for the unmodeled temporal variation of mass, so that the predicted worst-case flight conditions associated with flutter remain conservative to the true unstable conditions. The flutterometer could be used to determine safe test points for envelope expansion because the unstable flight conditions are predicted by use of data throughout the flight to account for unmodeled temporally varying dynamics.

This work was done by Rick Lind and Martin Brenner of Dryden Flight Research Center. For further information, access the Technical Support Package (TSP) free on-line at www.techbriefs.com under the category.

This invention is owned by NASA, and a patent application has been filed. Inquiries concerning nonexclusive or exclusive license for its commercial development should be addressed to the Patent Counsel, Dryden Flight Research Center [see page 20]. Refer to DRC-98-01.