Development of a bridge live load model for serviceability limit states

Document Type

Conference Proceeding

Date of Original Version



In 1994, the American Association of State Highway Transportation Officials (AASHTO) introduced the Load and Resistance Factor Design (LRFD) Bridge Specifications. The LRFD, based on a probabilistic design philosophy, was calibrated based only on the ultimate strength limit state (ULS). Other limit sates such as serviceability and fatigue were determined according to the performance of existing bridges designed using previous design codes. Live load deflection limits, historically carried over from previous code editions dating back to the 1930s, are now made optional (except for orthotropic decks). Such limits were first introduced in response to complaints of objectionable vibration of some bridges. In an investigation by the Bureau of Public Roads, bridges with objectionable vibrations, subjectively determined by human tolerance levels, were linked with deflections in excess of Span/800. Although not strictly enforced by the LRFD, many state highway departments specify their own deflection limits based on a rather arbitrary criteria. Wisconsin, for example, has a relatively low legal limit of 80 kips (356 kN) yet enforces one of the strictest deflection limits of Span/1600 whereas Michigan with the highest legal load tolerance of 164 kips (730 kN) has a deflection limit of Span/800. In addition, the live load used in the calculation of girder displacement varies considerably from one state to another. For example, Idaho uses the HS20 truck with dynamic load allowance, whereas California uses the HS20 truck plus lane load plus the dynamic load allowance. The use of load factors is another source of variation among state design practices; Arkansas uses factored live loads while Colorado does not for the same live load type. These variations among states can ultimately lead to non-uniformity in design and result in a wide range of girder stiffness for the same load demand. The objective of this study is to present results of a reliability-based analysis of live load deflections in steel girder bridges. The analysis uses actual truck traffic load data to calculate maximum deflections. Girder displacement values are forecast for different return periods (i.e. 1-, 2- ⋯ 75-years) using extreme value theory (EVT). Results presented herein provide an alternate approach for developing a deflection limit state. © 2008 ASCE.

Publication Title

Proceedings of the 2008 Structures Congress - Structures Congress 2008: Crossing the Borders