For as long as people have been making things, engineers have worked on improving how those things are made, often by designing materials that are stronger, more resilient, more efficient and less costly to produce.
Understanding the mechanical properties of advanced materials―novel materials with unique or enhanced properties―has taken on new manufacturing importance because advanced materials, usually at the nanoscale or smaller, have the potential to drive major improvements in consumer products such as device screens and computer chips, and in science and health in areas such as air travel, space exploration, solar energy production and medical devices.
“When looking at the structure of materials, you have different pieces that come together and join or adhere together or arrange themselves, like building blocks,” says Ganesh Balasubramanian, an associate professor of mechanical engineering and mechanics whose lab focuses on understanding the mechanical properties of advanced materials through computational and experimental methods. “For materials at the atomic scale, the technology is advanced enough that we can actually manipulate where atoms want to sit together.”
Changing those patterns can change the key properties of the material, such as its elasticity, or electrical and thermal conductivity.
Balasubramanian’s lab creates computational models to describe how the atoms fit together in order to address various design challenges. They do this by creating samples, testing the material and applying what they learn.
“However, if we keep doing these models for each and every type of material that we come across, it will just take ages,” says Balasubramanian.
So Balasubramanian and his team have added a new and potentially groundbreaking step: predictive engineering. Prediction leverages data and advanced computing to speed up the process exponentially.
“When the samples we create are tested and the experiments tell us the samples are good, we have achieved our objective,” says Balasubramanian. “If the experiments tell us they are not good, then the experiments can be used to reinstruct and inform the models.”
He and his team are primarily applying the predict-build-test-repeat approach to a very promising material class known as high-entropy alloys, a subset of multi-principal element alloys. This new class of materials are alloys formed by mixing significant and varying proportions of multiple elements. Preliminary studies have demonstrated that multi-principal element alloys have superior mechanical strength and hardness, making them ideal for turbine blades, medical implants, as a protective coating on components like ship surfaces and aerospace parts.
Balasubramanian, working with colleagues at Ames Laboratory, recently developed a hybrid version of an algorithm called the Cuckoo Search to overcome the limitations of current materials-simulation models and accelerate the computational modeling of complex alloys. The Cuckoo Search is inspired by the evolutionary strategy and brood parasitism of Cuckoo birds.
“In evolutionary biology, traits compete down the generation and the best traits outcompete those with worse traits, and create new generations,” says Balasubramanian. “That can describe the process of creating materials with the best traits, the best fabrication strategies. Using predictive methods such as the Cuckoo Search speeds up that process.”
In fact, in this most recent work, the predictive process sped search time for materials design 13,000-fold.
“That was indeed startling,” says Balasubramanian. “What took about a day to accomplish, can now be done in seconds.”
Ganesh Balasubramanian's research interests are in advanced energy and structural materials, nanoscale transport and mechanics, and predictive engineering. He received his Ph.D. from Virginia Tech and was a postdoctoral research associate in theoretical physical chemistry at TU Darmstadt in Germany. Balasubramanian is a 2020 recipient of an NSF CAREER Award.
