Quick look
Iowa State engineers are joining researchers from across the country to develop new tools to simulate electrochemical processes critical to energy, health care and manufacturing. The U.S. National Science Foundation is supporting the project.
AMES, Iowa - There isn't a good and fast way to simulate the electrochemical processes critical for energy production, energy storage, health care technology and advanced manufacturing.
Researchers working together from three universities across the United States - Iowa State University, Tufts University in Massachusetts and Stanford University in California - aim to change that with an effort to build computational modeling tools for those processes in liquids or gases.
Those electrochemical systems touch daily life in surprising ways, including clean water through desalination technology, cheaper energy through better battery components, and better health care through smaller and cheaper devices.
Researchers say making high-fidelity modeling easy, fast and sharable can save time from discovery to products benefitting all of us.
Consider, for example, a desalination system, said Baskar Ganapathysubramanian, a mechanical engineer, Iowa State's Joseph and Elizabeth Anderlik Professor in Engineering and an associate director of the university's Translational AI Center. The membrane's structure is complex and porous. When an electric field is applied, charged particles - in this case salt ions - are separated from water.
It sounds simple enough.
"But the problem here is there are too many competing things happening," Ganapathysubramanian said. "There's the chemistry at the surface and we can zoom in on small surface features. But fluid next to the surface is also affected."
Accounting for various scales in distance and time - from microscopic interfaces to large systems - is also a challenge, he said.
"Think of this project as building a car. The 'car' is the open-source software. Most people shouldn't have to build a car to get where they need to go. They should be able to drive it easily, safely and quickly to their destination. That's what we're delivering: a powerful, reliable vehicle for scientists and engineers to test ideas and scale breakthroughs."
A new, collaborative research grant will help researchers study those problems and develop a modeling solution called FASTEST, "Framework for Advanced Simulation of multiphaSe ElecTrochemical Systems."
The U.S. National Science Foundation (NSF) is supporting the research with five-year grants: $1.15 million to Iowa State, $1.12 million to Tufts and $1 million to Stanford.
Project leaders are:
- Iowa State, Ganapathysubramanian and Adarsh Krishnamurthy, a professor of mechanical engineering and an associate director of the Translational AI Center
- Tufts, Hari Sundar, the Ada Lovelace Associate Professor in Computer Science; and Jeff Foster, a professor and the chair of computer science
- Stanford, Ali Mani, a professor of mechanical engineerin
Challenges and expertise
The researchers will work to develop "scalable, robust, and accurate simulation capabilities for the complex, multiscale dynamics of multiphase electrochemical systems," according to a project summary. Those simulations should also work across applications, including energy, medicine and manufacturing.
"Think of this project as building a car," Ganapathysubramanian said. "The 'car' is the open-source software. Most people shouldn't have to build a car to get where they need to go. They should be able to drive it easily, safely and quickly to their destination. That's what we're delivering: a powerful, reliable vehicle for scientists and engineers to test ideas and scale breakthroughs."
The researchers face three main challenges: building simulation technology that still works quickly when researchers adjust parameters and variables; creating a framework that doesn't depend on computing architectures such as central processing units or graphics processing units; and developing AI-ready tools that work quickly with large, complex datasets.
Additional problems include modeling the complex geometries in some applications, Krishnamurthy said. In the case of a dialysis machine, for example, modeling the geometries is difficult using existing methods.
"Addressing these challenges can significantly enhance the performance, efficiency, and affordability of critical technologies, particularly in energy production and storage," the researchers wrote.
The collaborating researchers from across the country "have so much diverse expertise," Krishnamurthy said, noting experts in applied math, computer programing languages, high performance computing, chemical engineering and fluid mechanics.
That's the kind of team that can only be assembled with the support of a federal agency, Ganapathysubramanian said. The NSF's Cyberinfrastructure for Sustained Scientific Innovation program ensures the team's work remains public, transparent and reusable, maximizing returns for taxpayers and enabling broad societal impact.
