MAE Seminar: Dislocation Avalanching in the Thermally Activated Regime

Title:

Dislocation Avalanching in the Thermally Activated Regime

Abstract: 

A nonequilibrium statistical thermodynamics treatment of thermally activated dislocation ensembles is presented that is consistent with internal state variable (ISV) theory based on nonequilibrium trajectories of grains/phases along a sequence of constrained local equilibrium states. Thermodynamic state relations based on constrained local equilibrium are used to define both configurational and thermal entropy changes associated with pending reactions in the energy landscape, directly linked to the probability of dislocation-barrier reactions. Considering individual grains or phases as subsystems, we define the degree-of-correlation of reactions in terms of the ratio of dissipation-weighted effective enthalpy to the maximum barrier enthalpy within each subsystem and for the ensemble comprised of all subsystems. Implications for scaling relations of intermittent flow and self-organized criticality of dislocation ensembles are considered that related to maximal intrinsic entropy production rate of state transitions and entropy change of pending reactions for the set of constrained local equilibrium states. We identify linkages between maximal entropy change of pending reactions and the principle of constrained maximal dissipation introduced by Ziegler. It is suggested that additional consideration of correlation of reactions can offer utility in applying physics-informed machine learning methods to achieve thermodynamically consistent reduced order crystal plasticity model forms.

Bio: 

Dave McDowell joined Georgia Tech in 1983 and is Regents’ Professor Emeritus in the GWW School of Mechanical Engineering. He served as Director of the Mechanical Properties Research Laboratory from 1992-2012 and as founding Executive Director of the Institute for Materials (now called the Institute for Matter and Systems) from 2012-2020, a Georgia Tech interdisciplinary research institute charged with developing and maintaining the campus materials user facilities and cultivating a campus-wide materials innovation ecosystem for research and education, including the incorporation of digital data science in materials research and education.

McDowell’s current research interests focus on microstructure-sensitive computational approaches to variability in fatigue of advanced alloy systems, including extreme value responses such as high cycle fatigue, novel concurrent atomistic-continuum (CAC) coarse-grained atomistic modeling for predictive materials simulation, multiscale chemo-physics modeling of point and line defect interactions with application to environmental effects, nonequilibrium statistical mechanics of hierarchically structured materials with dislocations, and hierarchical continuum multiscale modeling approaches including uncertainty quantification and propagation across length and time scales (cf. Wang, Y., Tran, A.V., and McDowell, D.L., Fundamentals of Uncertainty Quantification for Engineers: Methods and Models, 1st Edition, Elsevier, 2025; Uncertainty in Multiscale Materials Modeling, Eds. Y. Wang and D.L. McDowell, Elsevier, 2020, ISBN: 9780081029411). He has pursed development of methods that employ computational materials science and mechanics to inform design of materials, having co-authored a related textbook (Integrated Design of Multiscale, Multifunctional Materials and Products, Elsevier, 2010, ISBN-13: 978-1-85617-662-0). McDowell currently is a member of the editorial board of npj Computational Materials as well as several other journals, and served as co-Editor of the International Journal of Fatigue from 2008 to 2020. In 2019-2020, he was awarded the Georgia Tech Class of 1934 Distinguished Professor Award and was elected as a Fellow of TMS. He was awarded the Paul C. Paris Gold Medal in 2023 at the International Congress on Fracture, and is the 2023 recipient of the ASME Worcester Reed Warner Medal for outstanding contributions to the permanent literature of engineering.