Application of fracture mechanics to steel moment connections

Changbin Joh, Purdue University

Abstract

The 6.8 magnitude Northridge earthquake that shook California's San Fernando Valley on January 17 in 1994, did not cause the collapses of major steel structures, but steel moment connections, confidently designed and constructed in the past with traditional code simplification and common site welding techniques, were discovered not to meet our expectations. The main purpose of the present study is to show the applicability of fracture mechanics to the analysis of the steel moment connection failures during the Northridge earthquake. Other purposes are to develop the numerical procedure of determining the fracture strength of the steel moment connection using linear elastic fracture mechanics, and the probabilistic analyses of brittle fractures of connections. The steel moment connection failures during the 1994 Northridge earthquake and the design philosophy are reviewed and the post-Northridge earthquake experimental and analytical researches are examined. Possible causes of the steel moment connection failures are categorized into welding-related factors, design-related factors, and material-related factors. For the analyses, the idealizations of the steel moment connections considering each factor are studied. The stress concentration and the sate of stress of the connection are investigated using an idealized five-plate moment connection model. The brittle fracture strength of the welded-flange bolted web connection with backing bars is studied using linear elastic fracture mechanics. Post-Northridge connection tests are simulated using the finite element method. An edge crack is assumed at the column-weld interface to represent crack-like defects and other adverse effects. The energy release rate at the front of the edge crack at the failure load of the test is studied using the J-integral method. From post-Northridge connection test results and a finite element model, the effective fracture toughness of the column-weld interface is investigated. The effective fracture toughness is the maximum energy release rate of the material at the front of the crack at the failure load. A new method of calculating the brittle fracture strength of the connection is suggested using the effective fracture toughness. The validity of the suggested method is investigated through numerical examples. Based on the suggested method, the probability of brittle fracture of the connection is investigated using a sample weld defect distribution and a yield stress distribution. In addition, possible applications of this method are discussed.

Degree

Ph.D.

Advisors

Chen, Purdue University.

Subject Area

Civil engineering|Mechanical engineering|Mechanics|Urban planning|Area planning & development

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS