FlightGlobal.com
Home
Premium
Archive
Video
Images
Forum
Atlas
Blogs
Jobs
Shop
RSS
Email Newsletters
You are in:
Home
Aviation History
1951
1951 - 2280.PDF
rLIGHT, 16 November 1951 619 HIGH-SPEED FLUTTER A Review of Recent Investigations: I. E. Garrick's Anglo-American Lecture IN his paper entitled Some Research on High-SpeedFlutter, given at the Third Anglo-American Conferenceat Brighton recently, Mr. I. E. Garrick (Chief of the Dynamic Loads Division, Langley Aeronautical Laboratory) paid tribute to the pioneering, and continuing, British work on the flutter problem and mentioned in particular the classic lectures of Collar and Duncan. The flutter field, he recalled, was concerned with the study of circumstances whereby an aircraft or one of its components could deform, and thus absorb energy from a uniform air stream to an extent of damaging or destroying itself. Flutter, he said, might be regarded as a self-excited oscillation which occurred when the damping of a vibration mode (or combina- tion of modes) in flight became negative; hence, the manner of loss of damping in the approach to flutter was also signifi- cant. By contrast, buffeting was generally considered to be the response of the aircraft structure to flow disturbances. The cure or remedy for buffeting, if any, was generally achieved by proper treatment of the external flow conditions; for flutter the treatment was an internal one of change in the magnitude and distribution of damping, rigidity and mass of the structure. In the minds of many workers on flutter the suspicion was forming that at least a fraction of buffeting cases for near sonic speeds could be associated with loss of damping and nearness to flutter. Dealing with the role of theory in the attack on flutter, Mr. Garrick said that, since it was almost impossible to obtain experimental flutter information by changing only one variable at a time, theory would have to apply the missing links which allowed a unified presentation of the results of hybrid testing. Both the structural and the aerodynamic theoretical backgrounds were complicated. The structural side embraced methods of classical mechanical vibrations for non-classical configurations which had led, in recent times, to such recondite matters as non-self-adjoint partial differen- tial equations, which even the books of pure mathematics avoided. Linearized theory had been the mainstay of theoretical aerodynamics, despite all its limitations. Modern linearized theoretical aerodynamics was often regarded as a branch of theoretical acoustics, dealing with small distur- bances in motion, whether at subsonic or supersonic speeds. Even so, its application to general configurations, for vibration modes and speeds of interest, offered such com- plexity in practice that little applicable information existed. One of the chief simplifications in non-steady aerodynamics, he said, was the use of results of two-dimensional flow or strip analysis. This usage did not imply that three-dimen- sional effects were insignificant, but more often than not it was intended for unification of information and presentation rather than for the precise numerical content. The determi- nation of design flutter margins was hardly a task to be entrusted to linearized strip theory alone. Speaking of cantilever wing flutter for high speeds, the lecturer referred to experimental results for the flutter of uniform cantilever wings of zero sweep which had been obtained by several different techniques. He illustrated (Fig. 1) the trends of these results, together with the calcula- tions based on two-dimensional compressible flow theory. Both the experimental and calculated values had been divided by the calculated value for the incompressible case and were presented as the ordinate VjVo. The abscissa M was the Mach number of flutter. The curve for full span aspect ratio, AR=j, was obtained from wings of 9 per cent thickness, while the curve AR—i was from wings of 5 per cent thickness. The approximate locations of the section centres of gravity were at 45 per cent chord. Mr. Garrick then referred to the experimental curve which, he pointed out, was fairly flat at low Mach number and rose sharply at transonic speeds. This rise, he said, was probably associated with shift with Mach number in the dynamic centre of pressure. For wings of different aspect- ratios, centre-of-gravity positions, frequency ratios, and so on, the shape of the curve and its location might be quite different. Experimental results in the form V/Vo were of interest because they furnished a comparison, for a wide range of Mach numbers, with the simple calculations based on low-speed flow; they also indicated that, as long as .the phenomenon was governed by the chosen degrees of freedom in bending and torsion, a unification of the results could be achieved with the use of simple theory. The theoretical calculated speed for two-dimensional flow increased sharply at a Mach number of 1.3 and flutter solutions were not found above about M=x.4. The general shape of the theoretical curve enveloped the experimental data quite satisfactorily. Aspect ratio and thickness effects were expected to be particularly significant in the transonic and supersonic speed regimes. The results had general significance in that they indicated that the near sonic speed range was not necessarily a "flutter barrier." Instead, the indications were that beyond a certain near-sonic value the wing-flutter situation improved. Mr. Garrick then went on to present the curve AR=j in another manner which, he said, might produce a better grasp of the implications of the results for design criteria. The ordinate VjVo in Fig. 1 might be expressed as (V/d)l(VBld), where "a" was the speed of sound, or as M/N, where M was the stream Mach number and N an artificial Mach number, Vo la, which was based on the incompressible fluid two-dimensional flutter calculations. N was actually a function of the design parameters, torsional stiffness, centre of gravity, mass ratio (or altitude) and so on. The experi- mental curve of Fig. 1 plotted as M against N was shown in Fig. 2. Nearness to the 45 deg line was a measure of the nearness of the experimental results to the two-dimensional theory for the incompressible fluid which was the basis N. As the value of N was increased (for example, by increasing the wing torsional rigidity or by an increase in the operating altitude) a value of N was reached beyond which flutter of the type considered would not occur. Such a wing would not encounter wing flutter at any Mach number. 2-5- 2-0- t-5- 1O • 0-5- "••••• TWO AR=2 J V DIMENSI THEORY Aft=7 /I / DNAL 05 15 Fig. 1. Flutter of uniform-cantilever wings of zero sweep; experimental and theoretical. Fig. 2. Experimental trends of AR=7 from Fig. 1 plotted as M against N. SO 1OO r cent span) Fig. 3. Comparison of some calculations for a wing with concentrated weights. o=expen- mental points.
Sign up to
Flight Digital Magazine
Flight Print Magazine
Airline Business Magazine
E-newsletters
RSS
Events
