Artificial intelligence to help design future of composite manufacturing

Posted: 

Ohio State is teaming up with other universities to employ artificial intelligence to inverse engineer composite materials. 

AIM for Composites graphic

Through the Artificially Intelligent Manufacturing Paradigm for Composites, or AIM for Composites, the project’s mission is “to build an AI-enabled design approach for fundamental understanding and integrated material-manufacturing design of advance polymer composites.” 

“The goal is to try to utilize composites to make tools for stamping,” says ISE Department Chair Farhang Pourboghrat, a co-principal investigator on the project. 

It’s part of the Department of Energy’s $400 million allocation for research at Energy Frontier Research Centers. AIM for Composites is receiving a four-year, renewable $10.3 million grant. 

Ohio State’s ISE Department will work with Principal Investigator Srikanth Pilla of Clemson University, Brown University, University of Florida, Pacific Northwest National Laboratory, Savanna River National Laboratory and South Carolina State University on the project. 

Farhang Pourboghrat hi res
ISE Department Chair Farhang Pourboghrat

Pourboghrat says each of the higher education partners brings its own strengths, with Ohio State serving as the “pillar leader” for the modeling capabilities. 

“One of our pillars is in manufacturing,” he says. “The other is in machine-learning and AI. Collaboration is a big component of this. The good thing about the collaboration is utilizing strengths from other universities. We think this is just the beginning of our collaboration.” 

The objectives are to embrace all stages of a composite component’s life cycle, leverage AI models and build cloud tools, and inform and implement new material and process designs. 

Descriptive models are intense and expensive, Pourboghrat says. The universities already have the models. He offers additive manufacturing as an example. “The output given is what we start with. AI captures the relationships between input and output. We do this in a time-efficient way.” 

He says the expense to design a program like this is largely incurred at the beginning of the process when training the artificial model “because we need a lot of data points. We can then determine or assess how good the model is and use another set of data to validate the model.” 

The data will be stored in the cloud where all the partners can access it and “make that better over time,” Pourboghrat says. 

According to the abstract, the investigators intend to address the “significant gaps between the performance, economic and environmental targets, and current design and manufacturing approaches.” 

“This project will develop AI-enabled, cloud-based, inverse design tools that can accelerate the discovery and manufacturing of new high-performance composite materials,” the abstract states. “If successful, we envision that the inverse design paradigm can be equally applied to other materials systems such as metals/alloys and ceramic matrix composites with applications in the automotive, aerospace, civil and defense industries.” 

Pourboghrat says the DOE requires that the grant be utilized to recruit and train underrepresented minorities among students and faculty as part of the program. The ISE Department is working with the College of Engineering’s Chief Diversity and Inclusion Officer Lisa Barclay on that aspect. 

In addition to training the future workforce, AIM for Composites will also provide self-reliance for the U.S. “The future is inevitably determined by AI cyber-security,” Pourboghrat says. “Dependence brings liabilities. We’re creating more security for the nation. It’s very timely to compete in that regard.” 

He says this is just the beginning of bringing composite research to Ohio State’s ISE Department and will help boost its national rankings. “I feel very fortunate that, within a year, we are receiving very important and influential funding. I feel really excited about the future.” 

 

Story by Nancy Richison

Category: Faculty