UNIVERSITY PARK, Pa. — Porous materials can be found everywhere from the concrete in our buildings to the bones in our bodies. In buildings, lightweight, high-strength porous materials help conserve energy, insulate acoustics and bear higher loads. In armored vehicles and airplanes, they reduce fuel consumption. Porous rocks are used to mitigate climate change by serving as a medium for carbon dioxide storage and geothermal heat extraction. Despite the proliferation and broad applications of porous materials, very little is understood about how cracks form inside their microstructure and how they eventually fail.
Yashar Mehmani, assistant professor in the John and Willie Leone Family Department of Energy and Mineral Engineering, received a $629,000 Faculty Early Career Development Program (CAREER) Award from the National Science Foundation (NSF) to pursue an integrated modeling, experimental and educational plan to improve the basic understanding of failures in porous materials and develop a more accurate computational framework to predict them.
“There is a gap in our understanding of how porous materials fail at microscopic scales,” said Mehmani. “Filling these gaps matters to science, because it can answer key questions about the targeted design of durable and high-performing materials for energy, defense and civil infrastructure. It also helps us manage more sustainably subsurface resources in storing and producing energy or disposing of carbon. A requisite to all this is the ability to predict deformation and failure using fast algorithms to make reliable and time-sensitive decisions.”
The research strives to close gaps in understanding within three identified areas: science, engineering and education.
According to Mehmani, despite technological achievements with X-rays that can scan the microstructure of porous materials in breathtaking detail, many researchers and engineers are caught in a Catch-22 regarding their use in computational tools.
“There are great algorithms out there that can take the intricate geometry measured by a microscope and perform calculations on it to make predictions,” said Mehmani “The problem is such calculations are either too expensive, preventing real-time or in-time decision making, or they lack sufficient accuracy to be entrusted with the kind of high-precision engineering needed in numerous applications. Affordable models end up discarding so much detail in the way of reducing computational cost that they produce predictions which are difficult to label as reliable — most come without error bars.”
The research team’s first aim is to develop faster, more accurate models that sidestep any non-essential calculations. The current, more affordable, models discard details arbitrarily, whereas Mehmani proposes to zero in on calculations where they are needed the most. Machine learning will be used to augment and accelerate calculations further.
The team then plans to create an open-source dataset to validate the models and identify knowledge gaps. Porous microstructures will be replicated with a 3D printer and scanned to map crack growth. A main objective is to curate high-quality datasets that the boarder research community can use and build on.
Additionally, the team will address the educational gap by increasing student literacy on computational mechanics of porous media both at the high school-level and the university-level. Open-source software, data and tutorials will be disseminated among students and experts in the field. In partnership with the Center for Science and the Schools at Penn State, the team also plans to work with a high school teacher over multiple years to produce a syllabus that stirs student interest towards STEM education.
“Broadening computational thinking skills and raising awareness of porous media mechanics are a first step in ensuring student engagement in problems that have societal and global impact in our collective future,” said Mehmani. “I’m excited to receive the NSF CAREER grant because it allows me to take a holistic approach towards realizing that step.”