Civil and Environmental Engineering

Surrogate models: A potential foundation for simultaneous structures/materials design

October 12, 2018

12:00 p.m. - 1:20 p.m.

Posner Hall, Mellon Auditorium

Lori Graham-Brady, Ph.D.
Professor
Department of Civil Engineering
Director, Modeling Complex Systems IGERT Program (MCS IGERT)
Associate Director, Hopkins Extreme Materials Institute (HEMI)

Abstract

Design of structural components is traditionally performed by first selecting a material and subsequently designing a structural configuration that meets the given requirements. With the onset of materials by design, there is future potential to engage in simultaneous design of materials and structures to maximize performance. The key step in this optimization will be to define an objective function that contains parameters that describe both the material (such as microstructural descriptors) and the structure (such as geometry).

While such an objective function can be formulated explicitly with multi-scale models, typically this approach would be computationally infeasible. Surrogate functions that provide a simplified representation of the material provide a much more efficient alternative. While surrogate functions provide a significantly more efficient path to represent the structural-scale behavior of materials, they can lead to a number of challenges. If the material is represented by a large number of microstructural parameters, the high dimensionality of the surrogate function requires many samples in order to build an accurate surrogate.

Furthermore, some micro-scale behavior, such as sudden damage, can lead to discontinuities in the surrogate function, which makes it difficult to interpolate or collocate the results. This seminar will describe a number of approaches to building surrogates, including cases in which the micro-scale model provides key response values and/or gradients of key response values. 

Bio

Lori Graham-Brady is professor and chair of the Civil Engineering Department at Johns Hopkins University, with secondary appointments in Mechanical Engineering and Materials Science & Engineering. Her research interests are in computational stochastic mechanics, multiscale modeling of materials with random microstructure, and the mechanics of failure under high-rate loading.

She is the associate director of the Hopkins Extreme Materials Institute and previous director of an NSF-funded IGERT training program with the theme of modeling complex systems. She has received a number of awards, including the Presidential Early Career Awards for Scientists and Engineers (PECASE), the Walter L. Huber Civil Engineering Research Prize, and the William H. Huggins Award for Excellence in Teaching.

Upcoming Events