Final Report Summary - ACTAGREEN (Aeroelasticity Control for Transportation And GREen ENergy)
The project ACTAGREEN focuses on two closely related engineering topics of importance to the EU economy and its society. The unrelenting increase in the length of modern suspended-span bridges makes them increasingly vulnerable to wind-induced vibrations and instabilities known respectively as buffeting and flutter (several bridge-construction projects with central span in excess of 1000m are currently being built or planned). In much the same way large wind turbines with power outputs in excess of 10MW and blade length well in excess of 100m are currently being investigated. As the blades of these machines increase in length, while also being constructed from lighter and more flexible materials, they too become susceptible to flutter and buffeting. In the case of large offshore wind farms, which may be exposed to highly unsteady aerodynamic loading, buffet suppression is especially important if this power generation means is to become widespread and economically attractive.
Wind-induced vibrations fall into four main categories: flutter, vortex induced vibrations, buffeting and torsional divergence [1]. Flutter is an instability due to negative damping effects that can occur in a wind of sufficiently high speed. In effect, small structural movements produce small changes in the aerodynamic force system, which in turn produces further small changes to the structural movements and so on. Under some circumstances this feedback mechanism becomes unstable. In specific engineering terms, this is a self-excited vibration (like wheel shimmy on cars, and weave and wobble on motorcycle [2]). Buffeting is another forced response and is caused by turbulence or other disturbances such as wakes in the incoming wind flow. Like vortex-induced vibrations, buffeting can be a significant source of fatigue damage that again undermines the serviceability of the structure. Vortex induced vibrations are vibrations induced by alternating pressure variations on the structure due to the periodic vortex wakes that shed naturally in the air stream - the so-called Karman vortex street. When the shedding frequency of the vortex street matches a natural frequency of the structure a resonant forced oscillation occurs called a vortex-induced vibration. This phenomenon is usually related to structural fatigue damage and can have a negative impact on serviceability of the structure. Torsional divergence is a non-oscillatory instability again brought about by high wind speeds, where the lift moment overcomes the structure's torsional rigidity and cause the structure to twist and lift into the wind.
The central focus of the proposed work was to seek common techniques for the analysis and suppression of wind-induced oscillations in large flexible civil engineering structures.
The limits and potential of leading and trailing edge flaps in suppressing aeroelastic instabilities has been investigated first. It was shown that, although the stabilization of the system can be obtained relatively easily, the closed-loop system loses robustness rapidly (i.e. it is very sensitive to uncertainty and errors in its parameters) for wind speeds beyond the (uncontrolled) torsional divergence speed. This is an important result that sets a qualitative limit and a benchmark for the performance of all controllers using leading-edge and trailing-edge winglets. In was also found that, in the case of long-span bridges, it is possible to significantly increase the structure’s aeroelastic limits with good robustness margins by fitting the deck with controllable flaps of about 30% of the main span’s total length [3].
Wind tunnel tests have been carried out to confirm numerical findings, and leading and trailing edge controlled flaps proved effective to stabilise bridge deck section model above its aeroelastic (flutter) limit [4].
The idea of using passive systems, i.e. not requiring energy supply, has also been explored and a novel system devised, the so-called Flap Mass Damper (FMD), which combines the favorable aerodynamic properties of the flaps with a driving (inertial) force provided by the a vibrating mass which is suspended within the bridge deck and is used as an actuator for the leading and trailing flaps [5].
Another result of the research is related to the application of the so-called “strip-theory”, i.e. the assumption that the aerodynamic forces on a given section depend only on the flow field at that section. This is a standard and widely accepted assumption when it comes to self-excited forces (flutter). The same assumption is often employed also for turbulence related (buffeting) forces, however in this case it is known that the assumption holds valid only when the length scale of turbulence is much larger than the width of the deck [6]. The present investigation showed that when considering sections with large aspect ratios (such as those typical of long-span suspension bridges) the “strip theory” can be used regardless the length scale of turbulence, i.e. even if the ratio of the length scale of turbulence to the deck width is not large. The results are relevant also for wings or rotor blades where the spanwise variation of the geometry is small over lengths [7].
For more information: matteo.massaro@unipd.it; david.limebeer@eng.ox.ac.uk; http://control.eng.ox.ac.uk/Actagreen
References
[1] Dowell, E.H. (2015) "A Modern Course in Aeroelasticity", 5th ed, Springer.
[2] Pacejka, H.B. (2012), “Tire and Vehicle Dynamics”, 3rd ed, Butterworth-Heinemann.
[3] Bakis, K.N. Massaro, M., Williams, M.S. Limebeer, D.J.N. (2016) “Aeroelastic Control of Long-Span Suspension Bridges with Controllable Winglets”, Structural Control and Health Monitoring, accepted for publication.
[4] Gouder, K., Zhao, X., Limebeer, D.J.N. Graham, J.M.R. (2015) “Experimental Aerodynamic Control of a Long-Span Suspension Bridge Section Using Leading- and Trailing-Edge Control Surfaces”, IEEE Transactions on Control Systems Technology
[5] Bakis, K.N. Massaro, M., Williams, M.S. Graham J.M.R. “Passive Control of Bridge Wind-Induced Instabilities by Tuned Mass Dampers and Movable Flaps”, Journal of Engineering Mechanics, under review.
[6] Li, S., Li, M., Liao, H. (2015), “The lift on an aerofoil in grid-generated turbulence”, J. Fluid
Mech., vol. 771, pp. 16-35.
[7] Massaro, M., Graham, J.M.R. (2015) “The effect of three-dimensionality on the aerodynamic admittance of thin sections in free stream turbulence”, Journal of Fluids and Structures, Vol: 57, Pages: 81-90.
Wind-induced vibrations fall into four main categories: flutter, vortex induced vibrations, buffeting and torsional divergence [1]. Flutter is an instability due to negative damping effects that can occur in a wind of sufficiently high speed. In effect, small structural movements produce small changes in the aerodynamic force system, which in turn produces further small changes to the structural movements and so on. Under some circumstances this feedback mechanism becomes unstable. In specific engineering terms, this is a self-excited vibration (like wheel shimmy on cars, and weave and wobble on motorcycle [2]). Buffeting is another forced response and is caused by turbulence or other disturbances such as wakes in the incoming wind flow. Like vortex-induced vibrations, buffeting can be a significant source of fatigue damage that again undermines the serviceability of the structure. Vortex induced vibrations are vibrations induced by alternating pressure variations on the structure due to the periodic vortex wakes that shed naturally in the air stream - the so-called Karman vortex street. When the shedding frequency of the vortex street matches a natural frequency of the structure a resonant forced oscillation occurs called a vortex-induced vibration. This phenomenon is usually related to structural fatigue damage and can have a negative impact on serviceability of the structure. Torsional divergence is a non-oscillatory instability again brought about by high wind speeds, where the lift moment overcomes the structure's torsional rigidity and cause the structure to twist and lift into the wind.
The central focus of the proposed work was to seek common techniques for the analysis and suppression of wind-induced oscillations in large flexible civil engineering structures.
The limits and potential of leading and trailing edge flaps in suppressing aeroelastic instabilities has been investigated first. It was shown that, although the stabilization of the system can be obtained relatively easily, the closed-loop system loses robustness rapidly (i.e. it is very sensitive to uncertainty and errors in its parameters) for wind speeds beyond the (uncontrolled) torsional divergence speed. This is an important result that sets a qualitative limit and a benchmark for the performance of all controllers using leading-edge and trailing-edge winglets. In was also found that, in the case of long-span bridges, it is possible to significantly increase the structure’s aeroelastic limits with good robustness margins by fitting the deck with controllable flaps of about 30% of the main span’s total length [3].
Wind tunnel tests have been carried out to confirm numerical findings, and leading and trailing edge controlled flaps proved effective to stabilise bridge deck section model above its aeroelastic (flutter) limit [4].
The idea of using passive systems, i.e. not requiring energy supply, has also been explored and a novel system devised, the so-called Flap Mass Damper (FMD), which combines the favorable aerodynamic properties of the flaps with a driving (inertial) force provided by the a vibrating mass which is suspended within the bridge deck and is used as an actuator for the leading and trailing flaps [5].
Another result of the research is related to the application of the so-called “strip-theory”, i.e. the assumption that the aerodynamic forces on a given section depend only on the flow field at that section. This is a standard and widely accepted assumption when it comes to self-excited forces (flutter). The same assumption is often employed also for turbulence related (buffeting) forces, however in this case it is known that the assumption holds valid only when the length scale of turbulence is much larger than the width of the deck [6]. The present investigation showed that when considering sections with large aspect ratios (such as those typical of long-span suspension bridges) the “strip theory” can be used regardless the length scale of turbulence, i.e. even if the ratio of the length scale of turbulence to the deck width is not large. The results are relevant also for wings or rotor blades where the spanwise variation of the geometry is small over lengths [7].
For more information: matteo.massaro@unipd.it; david.limebeer@eng.ox.ac.uk; http://control.eng.ox.ac.uk/Actagreen
References
[1] Dowell, E.H. (2015) "A Modern Course in Aeroelasticity", 5th ed, Springer.
[2] Pacejka, H.B. (2012), “Tire and Vehicle Dynamics”, 3rd ed, Butterworth-Heinemann.
[3] Bakis, K.N. Massaro, M., Williams, M.S. Limebeer, D.J.N. (2016) “Aeroelastic Control of Long-Span Suspension Bridges with Controllable Winglets”, Structural Control and Health Monitoring, accepted for publication.
[4] Gouder, K., Zhao, X., Limebeer, D.J.N. Graham, J.M.R. (2015) “Experimental Aerodynamic Control of a Long-Span Suspension Bridge Section Using Leading- and Trailing-Edge Control Surfaces”, IEEE Transactions on Control Systems Technology
[5] Bakis, K.N. Massaro, M., Williams, M.S. Graham J.M.R. “Passive Control of Bridge Wind-Induced Instabilities by Tuned Mass Dampers and Movable Flaps”, Journal of Engineering Mechanics, under review.
[6] Li, S., Li, M., Liao, H. (2015), “The lift on an aerofoil in grid-generated turbulence”, J. Fluid
Mech., vol. 771, pp. 16-35.
[7] Massaro, M., Graham, J.M.R. (2015) “The effect of three-dimensionality on the aerodynamic admittance of thin sections in free stream turbulence”, Journal of Fluids and Structures, Vol: 57, Pages: 81-90.