Shortening the learning curve: Schooling machine-learning algorithms to power materials digital twins and predictions

Presented by: Dr. Rémi Dingreville from Center for Integrated Nanotechnologies Sandia National Laboratories

Date: November 11, 2026

Time:  2:00 pm

Location:  Northlawn 2007

Abstract:  

Advanced computing is transforming how complex materials are designed, qualified, and deployed by enabling detailed predictions of their aging and performance under realistic operating conditions. Yet the underlying process–structure–propertyaging relationships are high-dimensional, nonlinear, and increasingly beyond the reach of traditional brute-force simulations. In this talk, I will discuss how machine learning (ML) strategies confront this complexity by “schooling” algorithms with materials science— encoding physics and domain expertise directly into their architectures and training data. Through case studies ranging from thermomechanical simulations of composites and alloys, to simulations of manufacturing processes, I will show how these “educated” models speed up simulation predictions, enable performance-by-design, and form the computational foundation for digital materials twins.

Bio:

Rémi is a Distinguished Member of the Technical Staff and staff scientist at the Center for Integrated Nanotechnologies. He received a Ph.D. in Mechanical Engineering from the Georgia Institute of Technology in Atlanta GA. After being a postdoctoral appointee at Sandia from 2007 to 2009, Remi was an Assistant Professor at New York University before rejoining Sandia in 2011. Remi’s expertise is at the intersection of computational materials and data sciences, specializing in the application of machinelearning (ML) algorithms to materials manufacturing and materials reliability. He has published over 150 peer-reviewed articles on these topics.

The University of Alabama     |     Lee J. Styslinger Jr. College of Engineering