Feb 1, 2019 | By Cameron
Mathematicians at New York University 3D printed 15 generations of algorithmically-bred wing designs to determine what shapes were best for flapping. When a wing flaps, vortices are generated at the leading edge and the tail edge of the wing and the interaction of those vortices determine the efficiency of the wing. To come up with the best design, the team looked to biological evolution.
"We can simulate biological evolution in the lab by generating a population of wings of different shapes, have them compete to achieve some desired objective, in this case, speed, and then have the best wings 'breed' to make related shapes that do even better," said Leif Ristroph, an assistant professor at NYU’s Courant Institute of Mathematical Sciences and the paper's senior author.
The experiments took place in NYU's Applied Math Lab, where the researchers started with 10 different 3D printed wings. Cavities were modeled into each wing that hold and dispense two fluorescent dyes: red in the front and green in the back. When the wing is flapped in the flow tank, the dyes reveal the eddies and vortices. For their tests, each wing was “raced” against the others by measuring their speed. On short wings, the vortex that forms on the leading edge crashes into the rear vortex, causing turbulence and slowing the wing. Longer wings diffuse the leading vortex before it reaches the rear edge, resulting in a faster wing.
Traits of the winning/faster wings were “bred” (combined) with the traits of other faster wings, creating 3D printed “daughter” wings that were then raced against each other. The process was repeated for 15 generations, yielding faster and faster wing offspring. "This 'survival of the fastest' process automatically discovers a quickest teardrop-shaped wing that most effectively manipulates the flows to generate thrust," explained Ristroph. "Further, because we explored a large variety of shapes in our study, we were also able to identify exactly what aspects of the shape were most responsible for the strong performance of the fastest wings."
Their data can be used to improve airfoils on planes as well as submarine designs, but Ristroph believes there’s potential for energy capturing, "We think this could be used, for example, to optimize the shape of a structure for harvesting the energy in water waves." 3D printing makes these kinds of iterative, generational designs incredibly affordable; the costs of fabricating these wings with subtractive manufacturing methods would have limited the team to only a couple iterations, drastically reducing the optimization that was achieved with multiple iterations.
Posted in 3D Printing Application
Maybe you also like:
- Scientists build 'Replicator' 3D printer that uses light to turn liquids into 3D objects in minutes
- 3D printing microstructures that can be moved with temperature or light
- Google Arts and Culture, Stratasys team up on 3D printing historical artifacts
- Huntington Ingalls Industries delivers first 3D printed metal part for US Navy aircraft carrier installation
- Iwakuni Marines 3D print engine ship kit for repairing fighter jets
- Ford made history with largest metal 3D printed automotive part for Ken Block’s Hoonitruck
- World's first fully 3D printed concept car is a tribute to David Bowie
- Engineers create 3D printed smart objects with 'embodied logic'
- Volvo installs 3D printed Living Seawall tiles in Sydney Harbour to combat ocean plastic pollution
- UMD engineers create smallest-ever 3D printed fluidic circuit element
- Nascar driver Brad Keselowski started a 3D printing company
A pretty good iterative process, but deeply flawed mimicry of flapping. Nothing in nature flaps in a vertical motion like this; natural wings have a singular root axis orientation that must be used to allow for optimal movement and proper low resistance return to power stroke motion. They would have been better off using models of actual fast bird wings and worked more on rotation dynamics than just one aspect of vortex reduction to derive speed optimization. IMHO.