FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1952
1952 - 0170.PDF
76 FLIGHT, 18 January 1952 FLUTTER . . . tion on a particular type of aircraft in order to assess the probable effect of the change of dynamic characteristics on the results of the flight tests. (v) Excitation of control circuit by electrodynamic exciter.—In this method an electrodynamic exciter of the moving-coil type, supplied by a variable frequency alternating current, was attached to the control circuit and aircraft structure. The particular advantages of this system were that force, amplitude and frequency were easily controlled in flight, and an almost instantaneous cut-off of exciting force was possible. It should be remembered, however, that electrical impedance of the exciter would affect the dynamical impedance of the control circuit to some extent, although it was thought that this effect would not seriously prejudice the method of excitation. (vi) Manual excitation by jerking the aircraft controls.—As dis cussed before, this method had been used with success both here and in the United States, and it was suggested that, despite some disadvantages, it was a satisfactory method of excitation for aircraft other than those fitted with powered flying controls. Experience had shown that the majority of relevant modes of a particular aircraft could be excited by jerking the aileron, elevator or rudder controls as appropriate. Fig. 8 showed records obtained by this method on a jet fighter aircraft. Undoubtedly in using the "jerk" there was some risk of failing to excite the critical flutter mode, but it was felt that the extreme simplicity of the method justified its adoption as a standard method of excitation. (vii) Manual excitation by oscillating the aircraft controls.—This method had been tried, but in the authors' view it was not very satisfactory. The frequency of excitation was limited by the human element to a maximum of approximately 5 cyc/sec. Furthermore, it appeared difficult for the pilot to vary the force applied to the controls without also altering the frequency. The tendency appeared to be to lower the frequency when endeavouring to increase the force, and therefore it was difficult for the pilot to control, with any certainty, either the frequency or the amplitude excited. The method could not be entirely dismissed, however, since it had been used with success in the United States during flutter tests on a large bomber aircraft. ll/l/vv oo«- f Fig. 8. Waveforms obtained from $tick "jerk" tests on a jet fighter aircraft. SOO 30O \ 400 EQUIVALENT MR SPEED (KT) None of the methods of excitation through the control circuit mentioned under the above headings was considered satisfactory for aircraft fitted with powered flying controls, because the dynamic characteristics of these controls were such that neither con tinuously forced oscillations having sufficiently high frequencies, nor sufficiently "sharp" jerks, could be transmitted from the input end to the output end of the circuits. The only systematic methods appeared to be excitation of the airframe to produce inertia and aerodynamic forces on the controls surfaces, or the use of inertia exciters on the controls surfaces themselves. If such methods were considered to be too complicated, the only alternative suggested by the authors was deliberately to fly the aircraft through "bumps" with increments of air speed and to take recordings of fhe resultant oscillations as in the general method. Such a technique, although better than no flutter tests at all, was obviously far from satis factory and the development of some simple and reliable method of excitation of powered flying controls was considered to be a major outstanding problem. Consideration of excitation methods should not be concluded without reference to a recent American proposal in which it was suggested that transient response of the aircraft structure could be excited by small rocket charges set in the surface of the structure. By altering the geometrical shape and burning characteristics of the charge, particular force/time functions could be applied. So far as was known, this method was at present being developed and no flight trials had yet been made. Mr. Kirkby suggested that simple vibration recording instru ments, such as the R.A.E. Mk II vibrographs, should be fitted when the control jerking method of excitation was being used. It was advisable to use three such instruments to record the vibration vertically in the outer wing, and vertically and laterally in the tail unit, to minimise the possibility of overlooking a mode in which the damping was low. If, during the subsequent flight tests, a condition of low damping was encountered which necessitated investigation before proceed ing to higher air speeds, it was strongly recommended that multi channel recording equipment be fitted at this juncture to establish the structural mode and control surface motion directly in flight. The alternative process of embarking on flutter calculations using the frequency recorded in flight and the appropriate ground resonance tests modes had led to much exhausting and fruitless effort in such situations in the past. Multi-channel electronic equipment made in miniature was at present being developed for use in small aircraft in the foregoing circumstances. The lecturer observed that, if the continuous excitation tech nique was being used it was worth while to fit multi-channel recording equipment from the start. The additional amount of work involved was probably small compared with the work involved in the fitting of the continuous excitation equipment, and it was possible to obtain more accurate measurements of damping in the different modes encountered in the frequency range. Considerable effort was at present being put into the development of equipment which would permit direct measurement of the amplitude of the vibration having the forcing frequency only, so that the superimposed vibrations originating from the power plant(s) and aerodynamic "hash" were not recorded. In the meantime much could be done to eliminate unwanted frequencies by the use of electronic filters. The results were analysed with a view to presenting curves of damping against air speed for the modes excited. The particular method of expressing the damping was not important, but it had been found convenient to use the damping ratio C/Co (=actual damping/critical damping), as found from the resonance or decay curves. It was unnecessary to plot damping curves for values of C/Co greater than 0.05, which experience had shown to be the value in a reasonably damped mode. The main difficulty that arose in trying to obtain reliable damping values from the flight test records was caused by the unwanted vibrations, some fairly continuous and some of a tran sient nature, which were forced by power-plant vibrations, aero dynamic "hash" and "bumps." As mentioned elsewhere in the paper, much could be done to alleviate these problems by forcing relatively large amplitudes and by electronic niters. If continuous excitation was being used, the problem was simplified somewhat, because the knowledge that the amplitude associated with the forcing frequency was constant for that particular recording enabled one to apply the usual graphical analytical methods. When the decay oscillation records were obtained the problem was more serious, particularly when the damping was compara tively high. If the damping was comparatively low, it was possible to average successive logarithmic decrement values from peak to peak in the decaying oscillation. It was probable, therefore, that greater accuracy was obtained when the damping was low than when the damping was high. As explained in the first part of the lecture, control-surface flutter was affected by equivalent air speed and Mach number, among other variables. For these two particular variables, it would therefore be ideally desirable to cover all combinations of E.A.S. and Mach number that could occur. The amount of flight testing required to approximate to this ideal condition was obviously prohibitive, and experience suggested that the best compromise was to cover the range of equivalent air speeds at the greatest height at which the maximum E.A.S. could be attained. Mr. Kirkby suggested that the range of air speeds should be divided into suitable increments, with smaller percentage incre ments as the high end of the range was approached. For example, the increments for an aircraft having a maximum E.A.S. of 500 kt might be arranged as follows : 150, 200, 250, 300, 350, 375, 400, 410, and then by 10 kt increments to 500 kt. The decision as to what increments should be used would obviously be influenced by the results of the pre-flight flutter calculations and by the results emerging from the flight measurements as the tests pro ceeded. The decision as to how many air-speed increments to cover in any one particular flight had also to be based on similar grounds, but to some extent the decision could be taken by the crew of the aircraft. If the aircraft was instrumented with electronic equip ment and carried an observer, it was possible to get an indication of the approach to flutter either from a monitoring system, or from some type of pen recorder on which the decay curves could be watched. The approach to a low damping condition could often be detected visibly by the response of the pilot's controls or the aircraft structure, after applying the "jerk." The crew should therefore be instructed not to go beyond certain speeds in each flight, and to stop the tests if the damping appeared to be falling
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
