Next week: ** SPECIAL LOCATION ** Le, 12-Mar-24, 12:20pm Central
1 view
Skip to first unread message
Ellad Tadmor
Mar 5, 2024, 1:57:22 PMMar 5
to aem-solid...@umn.edu
AEM Mechanics Research Seminar
** NOTE: Due to technical difficulties in our regular venue, next week's Seminar will take place in Room 321 of the Mechanical Engineering building (ME 321) **
Tuesday 12-Mar-2024, 12:20pm Central
Prof. Jia-Liang Le
Department of Civil, Environmental, and Geo- Engineering, University of Minnesota
Title: Mechanistic Mapping of Random Fields for Stochastic Finite Element Simulations of Quasibrittle Fracture
Abstract: Continuum finite element (FE) modeling of damage and failure of quasibrittle structures suffers from the spurious mesh sensitivity due to strain localization. This issue has been investigated extensively for deterministic analysis through the development of localization limiters. This talk will present a mechanism-based model to mitigate the mesh sensitivity in stochastic FE simulations of quasibrittle fracture. The present model is formulated within the framework of continuum damage mechanics, and the spatial randomness of material properties is represented by homogenous random fields. Two localization parameters are introduced to describe the evolution of the damage pattern of finite elements. These parameters are used to guide the energy regularization of the constitutive law, as well as to determine the mapping of the random fields of material properties onto the finite element mesh. The model is applied to simulate the stochastic failure of quasibrittle structures of different geometries featuring different behaviors including both distributed and localized damage. It is shown that the existing local projection and local averaging mapping methods could lead to strong mesh dependence of the predicted mean and variance of the structural load capacity. To mitigate the spurious mesh sensitivity, the mapping of the random fields of material properties must be tied to the damage pattern, which may evolve during the loading process. This result has important implications for the recent trends in the machine-learning approach for constitutive modeling of quasibrittle materials.
For more information, visit the AEM Mechanics Research Seminar website:
In-person: Room 321, Mechanical Engineering, University of Minnesota
Online: Webinar registration link
Department of Civil, Environmental, and Geo- Engineering, University of Minnesota
Title: Mechanistic Mapping of Random Fields for Stochastic Finite Element Simulations of Quasibrittle Fracture
Abstract: Continuum finite element (FE) modeling of damage and failure of quasibrittle structures suffers from the spurious mesh sensitivity due to strain localization. This issue has been investigated extensively for deterministic analysis through the development of localization limiters. This talk will present a mechanism-based model to mitigate the mesh sensitivity in stochastic FE simulations of quasibrittle fracture. The present model is formulated within the framework of continuum damage mechanics, and the spatial randomness of material properties is represented by homogenous random fields. Two localization parameters are introduced to describe the evolution of the damage pattern of finite elements. These parameters are used to guide the energy regularization of the constitutive law, as well as to determine the mapping of the random fields of material properties onto the finite element mesh. The model is applied to simulate the stochastic failure of quasibrittle structures of different geometries featuring different behaviors including both distributed and localized damage. It is shown that the existing local projection and local averaging mapping methods could lead to strong mesh dependence of the predicted mean and variance of the structural load capacity. To mitigate the spurious mesh sensitivity, the mapping of the random fields of material properties must be tied to the damage pattern, which may evolve during the loading process. This result has important implications for the recent trends in the machine-learning approach for constitutive modeling of quasibrittle materials.
For more information, visit the AEM Mechanics Research Seminar website:
---
Ellad B. Tadmor, Ph.D.
Russell J. Penrose Professor
Department of Aerospace Engineering and Mechanics
University of Minnesota
https://dept.aem.umn.edu/~tadmor/
Russell J. Penrose Professor
Department of Aerospace Engineering and Mechanics
University of Minnesota
https://dept.aem.umn.edu/~tadmor/
Reply all
Reply to author
Forward
0 new messages