Ian Castro 65 th Birthday Workshop, Southampton University, 28-29. 3. 12. Aerodynamic Control of Flexible Structures in the Natural Wind Mike Graham Department of Aeronautics, Imperial College London. [Acknowledgements to Mark Frederick, Kevin Gouder, David Hankin, Eric Kerrigan, David Limebeer, Bing Feng Ng, Xiaowei Zhao]
Long-Span Bridges (Tacoma Narrows, 1940, steady 42mph wind)
Wind Turbine (Jutland, 2010), probably very strong gust + fatigue damage.
Wind Turbine Rotor Blades Buffeting by incident turbulence, both the turbulence of the natural wind and that due to wakes of other turbines (wind park problem). Classical 2-dof. Flutter not so far a problem but increasingly long flexible blades are tending to reduce the critical flutter speed.
Horizontal and Vertical Axis Wind Turbine Rotors [HAWTs and VAWTs] 3 bladed, upwind HAWTs dominate. Typically now O(100m) dia, 80m/s tip speed, tip speed ratio ~ 4 20.
Porous Disc used to simulate the macroflow-field of a wind turbine. K range 0 4.
17m diameter, 50KW HAWT at the Rutherford- Appleton Lab wind test site. 2½ cycles of time history of surface pressure at an inboard section showing effects of wind shear, yaw misalignment and incident turbulence.
Wind Parks: wake impact
Development in ABL of u 2 (left) and U (right) at x/d = 2.0 and 7.0 [G.Hancock, Surrey Univ.]
Spectra of blade root pitching moment due to: (i) wind tunnel simulation of incident ABL turbulence (red), (ii) upstream turbine wake + ABL turbulence (blue).
Computational Methods. Free Wake Vortex Panel simulations of 2-bladed rotor flows (a) Single Rotor, (b) Rotor/Wake interaction
Indicial Response Method (convolution integral) Response, e.g.,, Impulsive response (VLM) of rotor Unit Response function System Response
Power and Thrust and Flapwise Blade Root Bending Moment Coefficient response to Velocity U(t). U(t) U(t)
Long-Span Bridges (Suspension or Cable- Stayed) Classical Flutter: Strong limitation on maximum possible span due to decreasing torsional stiffness. Buffeting (by turbulence in the natural wind): significant on many long-span bridges. Self-buffeting and VIV: deck-cross section design aims to minimise the effects.
Humber Bridge (UK) and Great Belt Bridge (Dk)
Humber Bridge Heave-Torsion Flutter (critical wind speed 58.4 m/s) Thin aerofoil approximations (a) frequency domain, (b) time domain
Eg. Surface panel method + free wake (time domain)
Forth (UK) and Akashi (Japan) bridges [Truss structures to provide deck torsion stiffening.]
Messina Straights: split deck solution. (to stiffen torsion and modify aerodynamics: wind tunnel test model)
Controlled Flap Solution. 2m trailing edge flap with moderate gain = 2.0*(1+i) * Deck Torsion Angle (critical wind speed is increased to 69.2m/s)
Rapid Actuation Gurney Flap : Disturbance Suppression Static Characteristics. Lift Effect Increase C L over incidence range Increase C L max Drag Effect Increase (L/D) max Increase in C D Improve Aerodynamic Characteristics Increase in stall incidence
Rapid Actuation Gurney Flap (water flume test)
Gurney Flap fluctuating load control
Alternative to Hinged Flaps: Adaptive Camber An adaptive camber wing has been developed with a skin-structure arrangement that seems to offer potential. A benchtop model and 0.5 span windtunnel model have been built and tested. Frequencies > 1hz are possible. (T. James, L. Iannucci, Aeronautics Dept., Imperial College).
Flutter and Buffet Control : Initial Challenge (working with control engineers) He still thinks winding numbers are used for counting rivets What are these old blokes worried about? We can sort this out in three months Teach him to spell Bisplinghoff and he thinks he s an aerodynamicist
Solution Format and Sign Conventions
Control System Configuration
Plain-Deck Response. (Flutter and Torsional Divergence)
With Controlled Trailing-Edge Flap
Wind Buffeting (Mixed Problem)
Truncated Continued fraction approximation of the von Kármán turbulence spectrum
Buffet Response to Simulated Turbulence (von Karman Spectrum, Thin Aerofoil Theory). (i) Bridge Deck section with controlled TE and LE flaps; (ii) Wind Turbine Blade Section with TE flap, LQG control: ~60% reduction in RMS achieved.
Forth Second Crossing Wind Tunnel Model fitted with Controlled Flaps.
theory... Conclusions Wind Turbine and Bridge stability and buffeting analysis carried out using classical unsteady aerodynamics re-cast in a control format. [Thin aerofoil theory appears satisfactory for determining flutter derivatives. Rational approximations used.] Significant increases in flutter speed are possible particularly via dual (leading and trailing) flap control. Trailing flap buffeting study shows that up to 80% reductions in buffet loads due to incident turbulence are possible on a wind-turbine blade or moderately streamlined deck section. Experimental work in progress, so far largely supports