Micromechanical Simulation of Earthquake-Induced Fractures in Welded Steel Structures

Stanford University, Stanford CA

Investigators

Abstract

9988902 Deierlein Characterized by high-strain low-cycle conditions, earthquake-induced fractures such as those observed in the 1994 Northridge and 1995 Kobe earthquakes are quite different from low-strain high-cycle fatigue fractures that have been extensively studied in bridges. Techniques to analyze and predict high-strain (inelastic) fracture are not nearly as well developed as those for other limit states. Thus, engineering researchers currently rely almost exclusively on empirical data from full-scale tests to develop fracture resistant welded connection details. The proposed research is to develop and apply computational methods for simulating inelastic earthquake-induced fractures with an emphasis on ductile crack initiation and growth. The scope includes complementary computational (finite element) and experimental studies that will utilize state-of-the-art micro-mechanical models to analyze ductile cracking and its potential for triggering brittle cleavage fracture. Detailed finite element analyses will include models that consider the interaction of plastic strains and stress triaxality that lead to void growth and coalescence. Results from the micro-mechanical models will be compared with conventional fracture indices (KI, J, CTOD), and physical tests will be used to both calibrate stress/strain and fracture toughness properties of base and weld metals and verify the simulations. Collaboration with researchers at the Nippon Steel Corporation and Chiba University will provide opportunities to study data from extensive Japanese research conducted in the aftermath of the Kobe earthquake. This project is supported under the 3rd-year competition under NSF 98-36. "US Japan Cooperative Research in Urban Earthquake Disaster Mitigation."