Abstract

A common configuration of decks on long span cable stayed bridges is to have local stiffness provided by plate girders located near the two edges. However, these sections are susceptible to vortex-induced oscillations that can be disturbing to bridge users and may potentially lead to other problems such as fatigue damage. A recent wind-tunnel investigation was conducted using a generic edge girder section model with a variable edge girder depth. This was done to capture the response characteristics of a range of practical cross-section geometries. The paper outlines the characteristics of both vertical and torsional vortex-induced responses observed in this investigation. Incorporating these response characteristics with the results of various wind tunnel studies, the effectiveness of baffle plates as possible mitigation measures is discussed. The characteristics of the responses, specifically the differences between the vertical and torsional vortex-induced oscillations, are important when optimizing the design of mitigation measures. Approximate prediction methods using Strouhal number estimates of basic sections, which allow specific mitigation measures to be incorporated early in the design process, are discussed. Guidelines for conducting detailed wind tunnel investigations on these sections are also proposed. VORTEX-INDUCED OSCILLATIONS Vortex-induced oscillation refers to the oscillatory motions of a structure caused by the regular shedding of vortices. This shedding induces fluctuating forces around the surface of the structure. Although vortex oscillations are self-limiting in amplitude, the magnitude of response is a critical aspect. Many design codes include criteria based on perceptible motions, whereby a structure is considered unserviceable if motions exceed a given displacement or acceleration. In BD 49/01 [BD 49/01], a guideline for assessing vortex excitation effects is given in terms of the vertical acceleration. Accelerations are considered acceptable up to approximately 2.5 mm/s, while pedestrian discomfort is expected for STRUCTURES 2006 Copyright ASCE 2006 Structures 2006 those over 40 mm/s. Using pedestrian discomfort as the limit state, these values are comparable to those presented by the American Society of Civil Engineers (ASCE), which limits vibrations due to vortex shedding to 5.0% g, where g is the acceleration due to gravity [ASCE]. Wind tunnel section model testing is the primary method used to estimate the susceptibility of a bridge section to vortex-induced oscillations. Plate girder bridge sections are susceptible to excessive motions due to vortex excitation as the sharp corners of their geometries, shown in Figure 1, promote the formation of separation zones and subsequently vortices. These excitations generally occur at relatively low wind speeds, and typically excite the fundamental vertical and torsional modes of vibration. The motions have restricted amplitude that typically do not reach destructive levels, but can lead to potential fatigue problems and serviceability issues [Nakamura, 149-169]. Critical vortex induced responses of a bridge section can be examined in the wind tunnel on a section model, allowing mitigation measures to be incorporated in the section at the design stage. FIGURE 1 TYPICAL PLATE GIRDER BRIDGE CROSS-SECTION [KING] PREDICTION OF VORTEX-INDUCED OSCILLATION ISSUES The frequency of vortex shedding can be defined according to the Strouhal Number relationship. The Strouhal Number (St), specific to the cross section, relates the shedding frequency (f), the across-wind dimension of the structure (D), and the mean wind speed (U) as follows:

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call