In 2018, passengers aboard a flight to Australia experienced a terrible 10-second dive when a vortex that followed their plane crossed after another flight. The collision of these eddies, the airline suspected, caused violent turbulence that led to a free fall.
In order to develop aircraft that can maneuver better in extreme situations, researchers at Purdue University have developed a modeling approach that simulates the entire process of a vortex collision with reduced computing time. This physical knowledge could then be incorporated into design codes to make the aircraft respond appropriately.
The simulations that aircraft designers currently use capture only a subset of the vortex collision events and require extensive data processing on a supercomputer. Not being able to simply simulate everything that happens when vortices collide has limited aircraft designs.
With more realistic and complete simulations, engineers could design planes like warplanes that can perform more abrupt maneuvers or helicopters that can land more safely on aircraft carriers, the researchers said.
"Airplanes in extreme conditions cannot rely on simple modeling," said Carlo Scalo, an associate professor of mechanical engineering at Purdue with a courtesy appointment in aerospace.
"Just to fix some of these calculations you have to run them on a thousand processors for a month. The aircraft design needs a faster calculation."
Engineers would still need a supercomputer to run the model developed by Scalo's team, but they could simulate a vortex collision in about a tenth to a hundredth of the time with far fewer computational resources than would normally be required for large-scale calculations.
The researchers call the model a "Coherent Vorticity Preserving (CvP) Large Eddy Simulation (LES)". The four-year development of this model is in the Journal of Fluid Mechanics.
"The CvP-LES model is capable of capturing super-complex physics without having to wait a month on a supercomputer because it already contains knowledge of the physics that calculations would need to meticulously reproduce on an extreme scale," said Scalo.
Former Purdue postdoctoral fellow Jean-Baptiste Chapelier led the two-year process of creating the model. Xinran Zhao, another Purdue postdoctoral fellow, performed complex, large-scale calculations to prove the accuracy of the model. These calculations allowed the researchers to create a more detailed account of the problem with more than a billion points. For comparison, a 4K ultra high definition television uses approximately 8 million dots to display an image.
Building on these foundations, researchers applied the CvP-LES model to the collision events of two vortex tubes, known as clover-leaf vortices, which are known to follow the wings of an aircraft and "dance" as they reconnect.
This dance is extremely difficult to capture.
"When eddies collide there is a collision that creates a lot of turbulence. It is very difficult to computationally simulate because an intense localized event occurs between two structures that look pretty innocent and uneventful until they collide," Scalo said.
Using the Brown supercomputer in Purdue for medium-sized computation and Department of Defense facilities for large-scale computation, the team processed data on the thousands of events that occur when these vortices dance and built that physical knowledge into the model. Then they simulated the entire collision dance with their turbulence model.
Engineers could simply run the pre-built model to simulate vortices over any period of time to best resemble what happened around an airplane, Scalo said. Physicists could also downsize the model for experiments on fluid dynamics.
"The really smart thing about Dr. Scalo's approach is that he uses information about fluid dynamics to determine the best tactic for computing fluid dynamics," said Matthew Munson, fluid dynamics program manager for the Army Research Office, a member of Army Research US Army Combat Capabilities Development Command Laboratory.
"It's a smart strategy because it makes the solving method applicable to a wider variety of regimes than many other approaches. There is tremendous potential for this to have a real impact on the design of vehicle platforms and weapon systems that enable our soldiers to to be succesfull." complete their missions. "
The Scalo team will use Purdue's newest community cluster supercomputer, Bell, to continue studying complex eddy currents. The team is also working with the Department of Defense to apply the CvP-LES model to large-scale test cases involving rotary wing aircraft such as helicopters.
"If you can simulate the thousands of events in the river exactly as they come from a helicopter blade, you can create much more complex systems," said Scalo.
This work was supported by the Army Research Office's Young Investigator Program, Award W911NF-18-1-0045. The researchers also confirm support from the Rosen Center for Advanced Computing in Purdue and the Supercomputing Resource Center of the US Air Force Defense Research Laboratory as part of subproject ARONC00723015.