FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1952
1952 - 0169.PDF
FLIGHT, 18 January 1952 75 FLUTTER Clearance Tests in Flight: Part II of the R.Ae.S. Lecture by Broadbent and Kirkby IN introducing the second part of the lecture (the first part, given by Mr. Broadbent, was dealt with in last week's issue of Flight), Mr. W. T. Kirkby said that the develop ment of the theoretical approach to flutter problems had been seriously hampered by the fact that little experimental full- scale evidence had been obtained to substantiate the theoreti cal work. Much of the evidence that had been obtained had been of a negative nature, i.e. no flutter had been predicted and no flutter incident had occurred. In the cases where flutter had occurred, it was often in degrees of freedom not considered in the pre-flight calculations, and had not always lent itself to theoretical explanation. There were two main aspects of full-scale flutter experiments. First, experiments forming part of a general research programme intended to check the validity of the appropriate theoretical calculations, and also intended to fill some of the gaps in existing knowledge by supplying information on modes in flight, and Mach-number effects. Second, flight experiments to form part of the development flying programme on prototype aircraft, or aircraft to which changes had been made which might affect the flutter conditions (i.e., tip tanks, external stores, modifications to control circuits). In the space available it was not possible to cover both the fore going aspects and this part of the paper was concerned primarily with the flutter clearance tests made during the normal test and development flying period of a particular type of aircraft. In the past, many flutter incidents had occurred in the course of test flying, often when some other flight characteristic of the aircraft was being investigated. Such incidents had usually occurred when the aircraft had exceeded the speed attained in previous flights, although a few incidents had occurred when flying in bumpy weather, at a speed less than the previous maximum, or at higher altitudes. These incidents had rarely been catastrophic, although in some cases there had been structural damage, but in any case they had an adverse effect on the morale of the crew of the aircraft. Fortunately, the linearity of coefficients that had to be assumed to make the theoretical calculations manageable was not realised in practice, otherwise most of the incidents would have been divergent and consequently disastrous. It was clearly desirable to adopt a flight technique which would indicate the approach to a flutter condition and thus lessen the risks to the crew and the aircraft. It was appreciated that the aircraft manufacturers already had formidable flight-test pro grammes, and that any further additions were unwelcome, but it was suggested that flutter check-tests could substantially reduce the risk of an unforeseen incident occurring, often without much additional flying, and with simple instrumentation. Often it might be possible to carry out the flutter clearance tests in the course of flights forming part of the generaltest and development pro gramme. It was only in comparatively recent years that serious considera tion had been given to flight flutter testing of a routine, rather than of a research, nature and consequently there was a lack of the background of experience and period of development that was available in many other fields of flight testing. The methods that were now considered represented present views, which might well be modified as further experience was obtained and as develop ment proceeded. The principle underlying all test methods was to excite the flutter modes of the aircraft from a controllable energy source, beginning at an air speed well below the maximum air speed, or calculated critical flutter speed, and to measure the response of the aircraft to the known energy input at increasing air speeds. The air speed was increased in small steps until the maximum permis sible speed of the aircraft was achieved, or until the response to the input energy showed that a flutter condition was being approached. In the general case it was desirable to excite all modes of the aircraft to ensure that the flutter modes were not missed; in the particular case, attention might be focused on a mode which might be suspect from the flutter point of view. The lecturer said that two methods of exciting the aircraft modes were considered in the paper :— (i) Forced vibration.—A sinusoidal force of variable frequency was applied to the aircraft structure or controls, as in ground resonance testing. (ii) Transient response.—A manual "jerk" was applied to the controls of the aircraft which would excite transient response in several modes at the same time. The degree of forcing of each mode, resulting from the jerk, would depend on the shape of the force/time curve, and it was theoretically possible to excite all the relevant structural modes by this method. The response of the aircraft structure to the energy input at each increment of air speed was determined from the damping in the modes. The value of the damping might be obtained from the resonance curves or from the decay oscillations, depending on which method of excitation was used. This damping value, which included the aerodynamic and structural terms, might be plotted against the equivalent air speed and the approach to a flutter condition would be indicated by an adverse change of slope of the curve or approach of the damping to zero. The general method just outlined was not original. It had been developed from methods suggested by von Schlippe, who proposed measurement of amplitude response to continuous forcing, and Grossman, who advocated measurement of damping in decay oscillations. A preliminary investigation, in which the damping measured from decay oscillations excited by manual jerking methods was compared with theoretical damping values, was reported by Rosenbaum and Scanlan in 1948. The latter method had subsequently been used with success both in this country and the United States. In certain cases the manual "jerk" method was not satisfactory, and the general problem of excitation of the aircraft modes was now examined briefly. METHOD Of EXCITATION I THROUGH AIRCRAFT STRUCTURE THROUGH CONTROL CIRCUIT OR CONTROL SURFACE NON-MANUAL MANUAL I I I INERTIAL AUXILIARY INERTIAL SPRING ELECTRODTNAMIC "JERK" SUSTAINED I OSCILLATING AND OSCILLATOR i OSCILLATION AEROFOIL ECCENTRIC I I I (i) (ii) (.ii) (iv) (v) (vi) (vii) Fig. 7. "Tree" summary of various methods of excitation. Mr. Kirkby went on to say that experience had shown that control-surface flutter did not occur at frequencies higher than five to six times that of the lowest structural frequency, and it was not considered necessary to excite frequencies higher than this during the flutter clearance tests. The various methods of excitation which had been considered were summarized in Fig. 7. (i) and (iii) Inertial excitation of aircraft structure or control surfaces.—This method was used generally in ground resonance testing and had been adequately described elsewhere. Care had to be taken to ensure that the mass of the exciter and associated driving equipment did not alter appreciably the modes of the aircraft. If possible, the exciter should replace a piece of aircraft equipment of equivalent mass. The disadvantage of inertial exciters became apparent at very low frequencies, when an extreme ly large amount of out-of-balance was required to produce the desired force, and the weight of the exciter became prohibitive. For this reason inertial exciters would seldom be used to excite frequencies below approximately 3 eye/sec. When such exciters were fitted in control surfaces, care had to be taken to ensure that the control-surface mass-balance and inertia conditions were not seriously affected. (ii) Excitation of aircraft structure by oscillating auxiliary aero foil.—In this method an auxiliary aerofoil of svmmetrical section was attached externally to the aircraft structure and a sinusoidal change of incidence, of variable frequency, was forced through a linkage. So far as was known, this method had not been tried in flight, but a preliminary design study had been made at the R.A.E., from which it appeared that the scheme might be practicable for frequencies from 5 cyc/sec down to zero. The disadvantages of this method were that the force produced varied approximately with the square of the air speed, and also that it was difficult to position the aerofoil so that the airflow round the lifting and control surfaces of the aircraft was unaffected by the proximity of the aerofoil. Because of the complexity of the installation it was probable that this method was better suited to long-term research investigations than to tests of a routine nature. (iv) Excitation of control circuit by spring and eccentric.— Although attractively simple, this method of excitation suffered from the inherent disadvantage that the spring, in effect, linked the control circuit to the airframe and, consequently, the dynamic characteristics of the control circuit would be altered. Flutter calculation? would be necessary before using this method of excita-
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
