Professor Blair Perot of the Mechanical and Industrial Engineering Department is the principal investigator on a two-year, $1.18-million ARPA-E grant to support the creation of pioneering software for simulating offshore wind turbines. The new software will leverage an existing user animation interface called Blender to replace the currently used CAD software, which tends to have a steep learning curve, is commercial and expensive, and has a framework geared towards static views and design, not the complex, real-world dynamics of offshore wind turbines.
As Perot and his collaborators say, this Blender animation software “represents objects and the user experience in a way which is intuitive to dynamics and control simulations. This will vastly improve the user experience for the proposed software compared to traditional simulation approaches. The learning curve is much easier for Blender than it is for any CAD software.”
The researchers explain that “a floating off-shore wind turbine is a case study in coupled dynamics. Wave action can pitch and heave the platform. Aerodynamic drag and rotational inertia can pitch and yaw the platform. The solid structures can vary widely in their flexibility from the fairly stiff tower, to the partially flexible blades and trailing wing, to the very flexible foam pontoons.”
According to the research team, past practice has been to leverage physics simulation software off the CAD design software community and its interfaces. But some key issues with CAD simulation software are how difficult it is to use, its steep user learning curve, and the software maintenance/upgrade cycle.
By contrast, say Perot and his associates, Blender is an open source and free animation platform that has a fully developed, intuitive graphical user interface and that boasts a large pre-existing software user base in the movie and animation community. Blender was originally designed for the computationally demanding task of computer generated imagery and 3D computer animation for films.
“We will leverage this user interface for the floating offshore-designer’s particular needs,” the researchers explain. “In this way, our core platform framework will be maintained and continually upgraded by the far larger animation and game writing community. We propose to only develop and maintain the simulation physics and control components that are specific to the off-shore problem.”
The researchers observe that their project will develop software for the coupled simulation and control co-design of floating offshore wind turbines. In that context, the physics of the fully coupled solid, liquid, and gas motion will be solved using a particle-mesh algorithm called Affine Particle in Cell. Blender will serve as the user interface for this new co-simulation physics engine. Finally, the free Sandia software, Dakota, will then interface with the proposed co-simulation results through Blender to perform design optimization using a control co-design approach.
In addition, say the researchers, an interface to Matlab API will be provided through a Python script. The team’s proposed software will ultimately be open source, intuitive, extensible, fully supported, and have fully coupled dynamic modules written to be efficient on massively parallel supercomputers.
Finally, according to the research team, “Common offshore wind design components such as blades, towers, and platforms (e.g. spar buoys) will be readily accessible from a component library. Each component geometry is easily tailored within Blender. The modular nature of the software will allow additional computational modules besides simulation physics to be added.” (October 2019)