Bodony contributes compressible turbulence expertise to new PCI center
Daniel Bodony, an associate professor of aerospace engineering, recently joined the Parallel Computing Institute (PCI) faculty as an expert in the field of large scale parallel computing, with a focus on simulating compressible turbulence.
Bodony and his collaborators developed a code to solve scientific computing problems at the petascale level that made him a perfect fit for PCI’s new Center for Exascale Simulation of Plasma-Coupled Combustion (XPACC). PCI is a Coordinated Science Laboratory institute committed to advancing parallel computing technologies.Bill Gropp said.
XPACC is working to develop the computer science (CS) tools to bring scientific prediction codes to exascale computers. Exascale computers are the next generation of large parallel computers whose power demands are imposing hardware changes that dictate entirely new approaches to data structures, algorithms and scientific computing in general. The Center will apply these developments to the target application of creating a new method of managing combustion on the fundamental science level. The goal is to leverage exascale computers to create a path for cleaner, more efficient and more reliable combustion devices, such as aircraft jet engines.
Bodony’s main research code, which has features amenable to the envisioned CS advancements, will be the starting point for computational scientific developments and predictions in XPACC. He will be working with the other faculty members to implement their CS and algorithmic ideas into the original code, verify that new code works correctly and evaluate their performance while predicting plasma-coupled combustion.
“I’m the link between the pure computer science and scientific prediction,” Bodony said. “The algorithms we use now won’t fully take advantage of the performance of the planned heterogeneous exascale machines, so we’re working together to develop new ideas because it’s unclear how relevant our existing codes and expertise will be as software and hardware develops in the future. Our goal is to demonstrate on test bed machines that we have all the flexibility and performance and, with a reasonable assumption, that those would all scale up to an exascale machine.”
XPACC is looking specifically at the interaction of a compressible turbulent reacting flow with plasma. A compressible turbulent flow is characterized by significant unsteadiness and changes in the gas density, a motion that is commonly found in aerospace applications, such as near the wings of jets or within jet engines.
“We don’t understand this motion very well and we’re interested in manipulating it to reduce drag, reduce noise and increase engine efficiency,” Bodony said. “We don’t currently understand the basic mechanics to do simple models, so we have to run really large simulations. Models are then developed based off those results.”
He added that they are also looking into the cause and effect of factors such as drag and noise.
“We’re realizing the problem is more complex than originally assumed, so we look at commonly assumed factors and evaluate whether they’re true or false and figure out why,” Bodony said.
Bodony joined the Illinois Aerospace department in 2006 and he received his B.S. and M.S. degrees from the School of Aeronautics & Astronautics at Purdue University and his Ph.D. in Aeronautics & Astronautics from Stanford University in 2005. After working at the NASA Ames/Stanford Center for Turbulence Research, he became in Illinois faculty member and holds an affiliate positions in the Mechanical Sciences and Engineering department and at NCSA.
Additionally, Bodony is a National Defense Science and Engineering Graduate fellow, an AFRL summer graduate fellow, is an ARCS fellow, a senior member of the AIAA on the Fluid Dynamics TC and a member of the American Physical Society. He recently received an NSF CAREER award in 2012 for his work on fluid dynamics.