Self-Adapting or Morphing Wind Turbine Blades | Paul Bomke

Flying and swimming animals live in unsteady environments. They are subject to atmospheric small scale turbulence, to highly unsteady flows in rivers or to vortex fields created by other animals while flying or swimming in schools, flocks or formations. Nevertheless, these organisms fly and swim with only minor de-viations from their paths. In addition, many animals that are moving in air or water possess highly efficient locomotory systems which at the same time provide an enormous degree of control over a wide range of situations. Such properties are highly desirable in technical fluid-based locomotion systems (e.g. planes, ships) and flow energy harvesting systems (e.g. hydro-and wind-turbines).

The most prominent feature of our biological archetypes is flexibility and, as a result, a high degree of agility and shape control. In many animals, for example in fishes, this is achieved by combining a flexible spine with segmented lateral muscle bands which can be activated locally along their length and provide a high level of control.Technical structures, in contrast, follow a very different mechanical paradigm: they are usually rigid.

Soft and flexible robotics are an exiting new field with a lot of potential in fluid dynamics, bringing us closer to mimicking the features and performance of living organisms.
In my project I am developing a prototype of a wing following the skeleton-muscle approach and a control method that changes the wing's shape based on the surrounding flow pattern. The main challenge is the use of dielectric elastomers, a type of electro-active polymers, for actuation of the wing. These actuators offer a high degree of flexibility as well as high energy efficiency but are fragile and require very high voltages.

The main parts of the project are

• The wing prototype: This is all about the mechanics and the actuation of the wing prototype. I'm creating a computational model of the wing to optimize each of the skeletons segments for maximum actuation while still being able to provide sufficient torque to handle the aerodynamic forces.
• Power supply: Dielectric elastomer actuators operate in a kilo volt range. In order to control a wing with multiple segments I'm developing a high voltage power supply with 20 channels, each with proportional voltage control of up to 4 kV.
• Machine learning control: An articulate wing with many degrees of freedom requires a sophisticated control algorithm. I'm making use of Reinforcement Learning to train a neural network to find adequate actuator control values in response to the pressure measured around the wing.

As a result, my project will provide a holistic picture of the challenges, benefits and downsides of dielectric elastomer actuators for wing structures and show the possible range of application within an engineering context.