Shougo Yabu

Investigation of the Combined Buckling Strength Formula for Corrosion-Damaged Compression Diagonal Members in Steel Truss Bridges

Eiji Iwasaki

 The deterioration of bridges constructed during Japan’s period of rapid economic growth has progressed significantly, and the proportion of bridges exceeding 50 years of service is expected to increase further. Under constraints of limited financial and human resources, efficient maintenance strategies are urgently required. In this context, establishing a rational and practical method for evaluating the structural performance of damaged members is of great importance.
 Compression diagonal members in steel truss bridges typically consist of box-shaped cross sections fabricated by fillet welding. Long-term deterioration can cause corrosion-induced section loss at welded joints, leading to weld separation referred to as corrosion-induced separation. Such damage reduces the local buckling strength of plate elements and influences the overall buckling behavior of the member. Accurate evaluation of coupled buckling behavior, involving interaction between global column buckling and local plate buckling, is therefore essential for structural safety assessment.
 Previous studies proposed a modified multiplicative formula to estimate buckling strength without elastic–plastic finite displacement analysis (FEA). However, the method was validated using only a limited number of member configurations, and its applicability to diverse bridge conditions remains unclear. In this study, elastic–plastic finite displacement analyses were conducted on 17 diagonal members with a single corrosion-induced separation, including five additional members derived from actual bridge specifications. Comparison between FEA results and predictions from the modified multiplicative formula showed errors ranging from −0.12σY to 0.10σY, confirming its general validity.
 To further reduce discrepancies, a correction formula using a coefficient Q′ was developed, where the variable was defined as the ratio of the averaged local buckling strength weighted by sectional area to that of the intact condition. The proposed equation reduced the error range to −0.12σY to 0.06σY, demonstrating improved accuracy and reduced unsafe-side error. Although accuracy was improved, residual errors remain, and additional data covering wider parameter ranges are required to enhance applicability.