A DCAMM seminar will be presented by
\n
\n
Professor Dr. Laura De Lorenzis
\nDept. of Mechanical and Processing Engineering
\nETH Zürich, Switzerland
\n
\n
\n
\n
Abstract:
\nWe propose a new approach for data-driven automated discovery of constitutive laws in continuum mechanics. The approach is unsupervised, i.e., it requires no stress data but only displacement and global force data, which can be realistically obtained from mechanical testing and digital image or volume correlation techniques; it can deliver either interpretable models, i.e., models that are embodied by parsimonious mathematical expressions, or black-box models encoded in artificial neural networks; it is one-shot, i.e., discovery only needs one experiment in principle - but can use more if available. The machine learning tool which enables discovery is sparse regression, leading to the automatic selection of a few relevant entries from a potentially very large model space.
\nAfter discussing the basics of the methodology, the talk illustrates its first applications to hyperelasticity, plasticity and viscoelasticity, and its latest extension to generalized standard materials. The latter allows the discovery of the category of material behavior (elastic,viscous, plastic or any combinations thereof) along with discovery and calibration of the specific material model within this category, while guaranteeing by construction material stability and thermodynamic consistency.
\n
\nDanish pastry, coffee and tea will be served 15 minutes before the seminar starts.
\n
\n
All interested persons are invited
\n X-ALT-DESC;FMTTYPE=text/html:A DCAMM seminar will be presented by
\n
\n
Professor Dr. Laura De Lorenzis
\nDept. of Mechanical and Processing Engineering
\nETH Zürich, Switzerland
\n
\n
\n
\n
Abstract:
\nWe propose a new approach for data-driven automated discovery of constitutive laws in continuum mechanics. The approach is unsupervised, i.e., it requires no stress data but only displacement and global force data, which can be realistically obtained from mechanical testing and digital image or volume correlation techniques; it can deliver either interpretable models, i.e., models that are embodied by parsimonious mathematical expressions, or black-box models encoded in artificial neural networks; it is one-shot, i.e., discovery only needs one experiment in principle - but can use more if available. The machine learning tool which enables discovery is sparse regression, leading to the automatic selection of a few relevant entries from a potentially very large model space.
\nAfter discussing the basics of the methodology, the talk illustrates its first applications to hyperelasticity, plasticity and viscoelasticity, and its latest extension to generalized standard materials. The latter allows the discovery of the category of material behavior (elastic,viscous, plastic or any combinations thereof) along with discovery and calibration of the specific material model within this category, while guaranteeing by construction material stability and thermodynamic consistency.
\n
\nDanish pastry, coffee and tea will be served 15 minutes before the seminar starts.
\n
\n
All interested persons are invited
\n URL:http://www.dcamm.dk/da/Kalender/2023/10/Seminar_No_770 DTSTAMP:20260919T191000Z UID:{E8212E5F-F661-4BED-A42A-434B7D10989E}-20231011T080000Z-20231011T080000Z LOCATION: Building 414, Room 016E, Technical University of Denmark END:VEVENT END:VCALENDAR