An Alternative to Rotary Wind Systems
‘Vibro-wind’ denotes the harvesting of energy from the wind as it flows around vibrating structures and is an emerging alternative to conventional rotary wind turbines. The basic science involves wind-induced vibration due to the non-linear fluid flow and vortices around flexible bodies and structures. Two key problems in this technology are: (1) how to convert wind energy into vibratory mechanical energy and (2) how to maximise mechanical energy conversion into electrical energy and storage from the vibration of a large array of hundreds of oscillators. A target application is for architectural facades in buildings, similar to, and as a complement to, solar energy panels.
By Francis C. Moon, Sibley School of Mechanical and Aerospace Engineering, Cornell University, USA
In addition to night-time generation of power in urban areas, vibro-wind devices can be effective in wind velocity environments as low as 2–3 metres per second (m/s), below typical rotary turbine start-up velocities of 9–10m/s. Thus, vibro-wind technology may have greater applicability than rotary systems in low wind environments.
Building-integrated power generation (BIPG) is an active area of architectural design. Its goal is to provide energy without a significant ecological footprint. Vibro-wind facade technology has an advantage over turbines on buildings because it avoids rotating dynamic loads on the structure as well as noise problems.
Physics of Vibro-Wind Power Generation
The flow of wind power P [Watts per square metre, W/m2] past an area ‘A’ normal to the flow velocity, V, is proportional to the air density and given by:
P = r V3 A/2
With the density of air of 1.2kg/m3, the power density of wind at V = 10m/s is 600W/m2. This density is the same for either rotary wind turbine systems or vibro-wind systems. One cannot hope to capture all of this energy. However, it might be possible to convert 30% of this power into structural vibration energy with a density of P = 180W/m2 (V = 10m/s). If one were to scavenge 30% of the structural vibration into electrical energy our figure of merit would be P = 54W/m2.
Commercial solar photovoltaic panels have an area power density of around 60W/m2. So the power output of vibro-wind panels on buildings might be comparable to solar photovoltaic technology. Although vibro-wind panels may be on the low end of solar panel power levels, the integrated energy may be comparable to or greater than solar, since wind may be available for 24 hours on a daily basis.
There are several modes of vibro-wind excitation including:
In the vortex shedding model there are two non-dimensional parameters: the Reynolds number, proportional to the velocity, and the Strouhal number, S = fD/U, that characterises the vortex frequency. Characteristics of fluid–structure dynamics have been summarised in the book by Blevins (1977), Fluid–Structure Vibration. For obstacles of the order of 50mm, and velocities of the order of 10m/s, the Reynolds number is around 30,000. In this regime, the alternating vortex flow behind a cylinder-type obstacle is well established with a given frequency. In this regime one can show that for 102 < Re < 105, S = 0.15 (where S = fsD/U) and fs is the vortex shedding frequency in cycles per second, then for an obstacle or flat plate of width D = 50mm, and U = 5m/s, fs= 15Hz. If the shedding frequency is in resonance with the oscillator frequency, a vibration amplitude of the order of 0.2 D is possible due to vortex shedding forces. For structural natural frequencies below the vortex shedding frequency, galloping vibrations can generate another self-excitation mechanism for structural oscillations. Typically, blunt-shaped bodies, such as cylinders with square cross-section, are most sensitive to vortex-induced wind forces as well as galloping wind forces. Also shape-edge structures, such as can be observed in so-called ‘stop-sign’ flutter, are susceptible to wind-induced vibrations. In vibro-wind devices, the design goal is to choose the most un-aerodynamic shape, which is in sharp contrast to applications to aircraft fluid dynamics design.
Prototype
As exploratory research, we have built arrays with four [2 x 2] oscillators of blunt-body cylinders in a 25 x 25cm wind tunnel in flow speeds in the range of 4–10m/s as well as a 25 [5 x 5] oscillator array for study in a larger wind tunnel (Figure 1). The oscillators were made of Styrofoam and had a natural frequency in the range of 8–10Hz. The larger array was tested in outdoor wind conditions. One of the blunt-body shapes chosen is the square cylinder 2 x 2cm with 6cm length attached to steel cantilevered ‘feeler gauges’. Piezoelectric benders were attached to each of the cantilever beams.
The galloping mode oscillator is a non-linear limit cycle instability that often exhibits a hysteretic ‘amplitude–wind speed’ behaviour. However, we were able to optimise the shape of the blunt body and the structural design of the composite piezo-beam to effectively eliminate the hysteresis, as shown in Figure 2. Here we have achieved a cut-in velocity below 2m/s, with large amplitude vibration equal to the diameter of the blunt body at wind tunnel speeds of around 4m/s2. The cantilevered oscillators were easily excited into vibrations at different angles of the wind and array plane. The lightweight Styrofoam cylinders were also easily excited in transient wind conditions and reached large amplitudes even in short-duration gusts.
Each oscillator was connected to a full rectifier bridge and the electrical charge from all the oscillators was stored in a capacitor. These preliminary experiments have been successful and have encouraged us to proceed with larger arrays. We have been able to store Joules of energy in a small capacitor, enough to light up LEDs for display. However, to date the efficiency of the piezo-rectifier system is not high. Further optimisation is required for commercial application (Figure 3a, b).
In Figure 3a the variation in wind velocity for outdoor experiments is compared with the voltage generated during steady wind tunnel tests. In Figure 3b, the charge stored in a capacitor is shown as a function of time for both the wind tunnel and outdoor tests. The outdoor tests show that the rectifier output will add charge to the capacitor in steps corresponding to the transient nature of the wind gusts.
Possible candidates for converting vibration energy to electrical energy are electromechanical, piezoelectric, piezo-polymer and electrostatic phenomena. Electromagnetic devices depend on coils and magnets, while piezoelectric devices are material property based. One key thrust of this project is the collection of electrical power from a large array of mechanical–electrical oscillators. Current research on vibration energy scavenging using piezo-materials has focused on the single oscillator problem. The Cornell research differs from those efforts in that we are using an array of oscillators with different phases. Our design goal has been to convert the AC electrical energy generated in the many different piezo-oscillators into DC electrical energy stored in either a capacitor or battery for use in building systems.
Architectural Applications
There is a great interest today in architectural building design in ‘building-integrated power generation’ or BIPG. In order for vibro-wind technology to be successful, the concept of placing vibrating structures on building facades must be acceptable to the professional architecture community. Fortunately there is precedent for dynamic elements in architecture originating in the kinetic sculpture or kinetic art world. In the 1990s the Japanese kinetic sculptor Susumu Shingu had installed large oscillating wind vanes on roofs, towers and domes that vibrated in the wind. More recently, the kinetic artist and designer Ned Kahn has designed panels on the scale of 2 x 12m with 80,000 vibrating plates for architectural projects in Charlotte, North Carolina, and Winterthur, Switzerland. These thousands of small vibrating plates are designed to produce visual wave-like effects as the wind blows around the buildings. Our goal will be to produce a vibro-wind technology that produces energy for internal building use as well as creating an aesthetic visual effect.
At Cornell University engineers are working with architecture students of Professor Kevin Pratt to design and fabricate architectural facades that can capture the wind and funnel the flow around hundreds of small vibrating structures with embedded piezoelectric devices to generate energy. One such design is shown in Figure 4. The openings serve to let in light as well as to create small wind tunnels to capture the wind.
The architectural issues that a vibro-wind energy system must address include the following:
There is emerging a new set of wind energy concepts based on vibratory motion and not rotary motion. These systems, while not as efficient as rotary machines, are more analogous to solar panel technology and have the potential for application to architectural design for building energy use, especially in urban areas where wind speeds are low and rotary turbines are not practical. The wind to vibration excitation mechanisms seem to be technically feasible and reliable. However, the key technology challenge is the transfer of energy from mechanical to electrical storage. Piezoelectric energy conversion shows promise but electromagnetic systems should not be ruled out at this time.
Acknowledgements
I would like to thank Professor Kevin Pratt, of Cornell Architecture, Cornell Center for Sustainable Energy. Cornell engineering and architecture students, Rona Banai, Albert Dodson, Zach Gould, Jamie Pelletier, Jared Valentin, Ranjeev Mahtani and Sang-tek Oh also made contributions and I would like to thank them too.
Biography of the Author
Francis Moon is a chaired professor of mechanical engineering at Cornell University in Ithaca, New York State. His field of expertise is applied dynamics of mechanical and electromechanical systems. He is an elected member of the US National Academy of Engineering. He is known for experimental work in non-linear and chaotic dynamics as well as magneto-mechanical systems such as magnetic levitation and superconducting systems. He is the author of several books on dynamics and magneto-mechanical systems and holds five patents. From 1987 to 1992 he was the Director of the School of Mechanical and Aerospace Engineering at Cornell. He has worked on the dynamics of fluid-elastic systems for 40 years. He was honoured by ASME in 2007 with their Lyapunov Award for lifetime achievements in non-linear dynamics. Recently he received an honorary doctorate from the Karlsruhe Institute of Technology in Germany.{/access}
‘Vibro-wind’ denotes the harvesting of energy from the wind as it flows around vibrating structures and is an emerging alternative to conventional rotary wind turbines. The basic science involves wind-induced vibration due to the non-linear fluid flow and vortices around flexible bodies and structures. Two key problems in this technology are: (1) how to convert wind energy into vibratory mechanical energy and (2) how to maximise mechanical energy conversion into electrical energy and storage from the vibration of a large array of hundreds of oscillators. A target application is for architectural facades in buildings, similar to, and as a complement to, solar energy panels.By Francis C. Moon, Sibley School of Mechanical and Aerospace Engineering, Cornell University, USA
{access view=!registered}Only logged in users can view the full text of the article.{/access}{access view=registered}‘Vibro-wind power’ is the harvesting of energy from the wind as it flows around commercial and residential buildings through the mechanism of vibrating structures. The basic science involves energy extraction from bodies induced to vibrate by the action of fluid flow and vortices around flexible structures. Our approach at Cornell University has been to consider the effects of wind on multiple interacting flexible structures, such as hundreds of small cantilevers mounted to a surface. Other vibro-wind concepts include large, fluttering wind-vane type structures, as well as flag or leaf and tree type flexible structures.
In our application the wind excites dozens up to thousands of small vibrating elements on panels attached to the structure (Figure 1), converting the kinetic energy into electrical energy that can be used in the operation of the building. There are two crucial steps in this process: one is the conversion of wind energy into vibration and the second is the conversion of mechanical vibratory kinetic energy into electrical energy. Our estimates of power output are comparable to solar panels and may complement solar panel systems during the night-time or serve as an alternative to solar panels for building applications, especially in urban areas.In addition to night-time generation of power in urban areas, vibro-wind devices can be effective in wind velocity environments as low as 2–3 metres per second (m/s), below typical rotary turbine start-up velocities of 9–10m/s. Thus, vibro-wind technology may have greater applicability than rotary systems in low wind environments.
Building-integrated power generation (BIPG) is an active area of architectural design. Its goal is to provide energy without a significant ecological footprint. Vibro-wind facade technology has an advantage over turbines on buildings because it avoids rotating dynamic loads on the structure as well as noise problems.
Physics of Vibro-Wind Power Generation
The flow of wind power P [Watts per square metre, W/m2] past an area ‘A’ normal to the flow velocity, V, is proportional to the air density and given by:
P = r V3 A/2
With the density of air of 1.2kg/m3, the power density of wind at V = 10m/s is 600W/m2. This density is the same for either rotary wind turbine systems or vibro-wind systems. One cannot hope to capture all of this energy. However, it might be possible to convert 30% of this power into structural vibration energy with a density of P = 180W/m2 (V = 10m/s). If one were to scavenge 30% of the structural vibration into electrical energy our figure of merit would be P = 54W/m2.
Commercial solar photovoltaic panels have an area power density of around 60W/m2. So the power output of vibro-wind panels on buildings might be comparable to solar photovoltaic technology. Although vibro-wind panels may be on the low end of solar panel power levels, the integrated energy may be comparable to or greater than solar, since wind may be available for 24 hours on a daily basis.
There are several modes of vibro-wind excitation including:
- Galloping vibrations (first analysed by the Dutch American, Den Hartog in 1932).
- Vortex-induced resonance (often referred to as von Karman vortex shedding).
- Bimodal flutter instability.
- Wind transient vibrations.
- Membrane wave-like vibrations; ‘flag flutter’.
In the vortex shedding model there are two non-dimensional parameters: the Reynolds number, proportional to the velocity, and the Strouhal number, S = fD/U, that characterises the vortex frequency. Characteristics of fluid–structure dynamics have been summarised in the book by Blevins (1977), Fluid–Structure Vibration. For obstacles of the order of 50mm, and velocities of the order of 10m/s, the Reynolds number is around 30,000. In this regime, the alternating vortex flow behind a cylinder-type obstacle is well established with a given frequency. In this regime one can show that for 102 < Re < 105, S = 0.15 (where S = fsD/U) and fs is the vortex shedding frequency in cycles per second, then for an obstacle or flat plate of width D = 50mm, and U = 5m/s, fs= 15Hz. If the shedding frequency is in resonance with the oscillator frequency, a vibration amplitude of the order of 0.2 D is possible due to vortex shedding forces. For structural natural frequencies below the vortex shedding frequency, galloping vibrations can generate another self-excitation mechanism for structural oscillations. Typically, blunt-shaped bodies, such as cylinders with square cross-section, are most sensitive to vortex-induced wind forces as well as galloping wind forces. Also shape-edge structures, such as can be observed in so-called ‘stop-sign’ flutter, are susceptible to wind-induced vibrations. In vibro-wind devices, the design goal is to choose the most un-aerodynamic shape, which is in sharp contrast to applications to aircraft fluid dynamics design.
Prototype
As exploratory research, we have built arrays with four [2 x 2] oscillators of blunt-body cylinders in a 25 x 25cm wind tunnel in flow speeds in the range of 4–10m/s as well as a 25 [5 x 5] oscillator array for study in a larger wind tunnel (Figure 1). The oscillators were made of Styrofoam and had a natural frequency in the range of 8–10Hz. The larger array was tested in outdoor wind conditions. One of the blunt-body shapes chosen is the square cylinder 2 x 2cm with 6cm length attached to steel cantilevered ‘feeler gauges’. Piezoelectric benders were attached to each of the cantilever beams.
The galloping mode oscillator is a non-linear limit cycle instability that often exhibits a hysteretic ‘amplitude–wind speed’ behaviour. However, we were able to optimise the shape of the blunt body and the structural design of the composite piezo-beam to effectively eliminate the hysteresis, as shown in Figure 2. Here we have achieved a cut-in velocity below 2m/s, with large amplitude vibration equal to the diameter of the blunt body at wind tunnel speeds of around 4m/s2. The cantilevered oscillators were easily excited into vibrations at different angles of the wind and array plane. The lightweight Styrofoam cylinders were also easily excited in transient wind conditions and reached large amplitudes even in short-duration gusts.
Each oscillator was connected to a full rectifier bridge and the electrical charge from all the oscillators was stored in a capacitor. These preliminary experiments have been successful and have encouraged us to proceed with larger arrays. We have been able to store Joules of energy in a small capacitor, enough to light up LEDs for display. However, to date the efficiency of the piezo-rectifier system is not high. Further optimisation is required for commercial application (Figure 3a, b).
In Figure 3a the variation in wind velocity for outdoor experiments is compared with the voltage generated during steady wind tunnel tests. In Figure 3b, the charge stored in a capacitor is shown as a function of time for both the wind tunnel and outdoor tests. The outdoor tests show that the rectifier output will add charge to the capacitor in steps corresponding to the transient nature of the wind gusts.
Possible candidates for converting vibration energy to electrical energy are electromechanical, piezoelectric, piezo-polymer and electrostatic phenomena. Electromagnetic devices depend on coils and magnets, while piezoelectric devices are material property based. One key thrust of this project is the collection of electrical power from a large array of mechanical–electrical oscillators. Current research on vibration energy scavenging using piezo-materials has focused on the single oscillator problem. The Cornell research differs from those efforts in that we are using an array of oscillators with different phases. Our design goal has been to convert the AC electrical energy generated in the many different piezo-oscillators into DC electrical energy stored in either a capacitor or battery for use in building systems.
Architectural Applications
There is a great interest today in architectural building design in ‘building-integrated power generation’ or BIPG. In order for vibro-wind technology to be successful, the concept of placing vibrating structures on building facades must be acceptable to the professional architecture community. Fortunately there is precedent for dynamic elements in architecture originating in the kinetic sculpture or kinetic art world. In the 1990s the Japanese kinetic sculptor Susumu Shingu had installed large oscillating wind vanes on roofs, towers and domes that vibrated in the wind. More recently, the kinetic artist and designer Ned Kahn has designed panels on the scale of 2 x 12m with 80,000 vibrating plates for architectural projects in Charlotte, North Carolina, and Winterthur, Switzerland. These thousands of small vibrating plates are designed to produce visual wave-like effects as the wind blows around the buildings. Our goal will be to produce a vibro-wind technology that produces energy for internal building use as well as creating an aesthetic visual effect.
At Cornell University engineers are working with architecture students of Professor Kevin Pratt to design and fabricate architectural facades that can capture the wind and funnel the flow around hundreds of small vibrating structures with embedded piezoelectric devices to generate energy. One such design is shown in Figure 4. The openings serve to let in light as well as to create small wind tunnels to capture the wind.
The architectural issues that a vibro-wind energy system must address include the following:
- compatibility with building facade design practices
- installation issues
- weather-resistant design
- maintainability
- vibro-wind panel lifetime
- fatigue of vibrating elements
- noise and acoustics
- aesthetic design
- integration into the building energy or grid system.
There is emerging a new set of wind energy concepts based on vibratory motion and not rotary motion. These systems, while not as efficient as rotary machines, are more analogous to solar panel technology and have the potential for application to architectural design for building energy use, especially in urban areas where wind speeds are low and rotary turbines are not practical. The wind to vibration excitation mechanisms seem to be technically feasible and reliable. However, the key technology challenge is the transfer of energy from mechanical to electrical storage. Piezoelectric energy conversion shows promise but electromagnetic systems should not be ruled out at this time.
Acknowledgements
I would like to thank Professor Kevin Pratt, of Cornell Architecture, Cornell Center for Sustainable Energy. Cornell engineering and architecture students, Rona Banai, Albert Dodson, Zach Gould, Jamie Pelletier, Jared Valentin, Ranjeev Mahtani and Sang-tek Oh also made contributions and I would like to thank them too.
Biography of the Author
Francis Moon is a chaired professor of mechanical engineering at Cornell University in Ithaca, New York State. His field of expertise is applied dynamics of mechanical and electromechanical systems. He is an elected member of the US National Academy of Engineering. He is known for experimental work in non-linear and chaotic dynamics as well as magneto-mechanical systems such as magnetic levitation and superconducting systems. He is the author of several books on dynamics and magneto-mechanical systems and holds five patents. From 1987 to 1992 he was the Director of the School of Mechanical and Aerospace Engineering at Cornell. He has worked on the dynamics of fluid-elastic systems for 40 years. He was honoured by ASME in 2007 with their Lyapunov Award for lifetime achievements in non-linear dynamics. Recently he received an honorary doctorate from the Karlsruhe Institute of Technology in Germany.{/access}
