Flange Local Buckling of High-Strength Steel I-Shaped Beams

Abstract

The use of high-strength steel in building and bridge construction can bring about many technological advantages. For example, it is possible to avoid the use of overly thick plates which often cause many problems in design, fabrication, transportation, and construction in the civil and architectural mega-structures. Slim members made of high-strength steel can enhance space availability, aesthetics, and freedom in design. Recently two types of high-strength steels with a nominal tensile strength of 800 MPa are developed in Korea. One is HSB800 steel for bridges and the other is HSA800 for building structures. Both steels are economically produced by thermo-mechanical control process without quenching and tempering process and can provide good weldability and notch toughness. The yield stress of these steels is 2 or 3 times higher than that of conventional mild steel. The elastic and inelastic flange local buckling of I-shaped beams fabricated from high-strength steel were studied in this paper. Flexural tests on full-scale I-shaped beams, built up from high-strength steels (HSB800 and HSA800) with a nominal tensile strength of 800 MPa and ordinary steel (SM490) for benchmark, were carried out to study the effect of flange slenderness on flexural strength and rotation capacity. Inelastic flange local buckling behavior was investigated by extrapolating current stability criteria (originally developed for ordinary steel) to high-strength steel. For flexural members that are expected to behave beyond the elastic region, more stringent slenderness ratio is required to control the buckling of high-strength steel. According to current code (AISC LRFD), the width-to-thickness ratio is limited in inversely proportion to the square root of the yield stress. For comparison purposes, specimens with ordinary steel (SM490) were also tested and showed sufficient flexural strength and rotation capacity in accordance with the AISC specification. The performance of high-strength steel specimens was also very satisfactory from the strength, but not from the rotation capacity, perspective. The inferior rotation capacity of high-strength steel beams was shown to be directly attributable to the absence of a distinct yield plateau and the high yield ratio of the material. From the stress-strain relationship of high-strength steel, analytical load-displacement relationships and moment-end rotation relationships of high-strength beam are obtained. As the analytical prediction, the ultimate rotation capacity of high-strength steel expected about 4~5 although the flange slenderness gets lower. The specimens with partial-height transverse stiffeners, or the specimens with no heat input to the beam bottom flange in welding the stiffeners, were able to exhibit the plastic rotation capacity required for plastic design. However, the specimens with full-height transverse stiffeners were welded to the tension flange in the plastic hinge region showed inferior plastic rotation capacity due to the brittle fracture of the heat-affected beam bottom flange at the stiffener location. The main reason for this brittle fracture was speculated to be caused by the heat input during stiffener welding. However, specimens with partial height stiffeners, or specimens with no heat input to the beam bottom flange, exhibited a plastic rotation capacity required for plastic design. Residual stress of SM490 and HSA800 was measured by instrumented indentation method. Residual stress measurements reconfirmed that the magnitude of the residual stress is almost independent of the yield stress of base metal. The mechanism of tensile fracture occurred in high-strength specimens was explained metallurgical aspect. It was confirmed that local buckling strength for slender flanges in current AISC design provisions is not suitable. The provision for flange local buckling of I-shaped beams with slender flanges is based on test of Johnson (1985). However, the test was conducted only to uniform moment beams and most of the specimens have noncompact or compact flanges and slender webs. For this reason, the provision for local buckling strength for slender flanges is not suitable, especially when the beam is subjected to linear moment gradient. New formula for flange local buckling strength of I-shaped beams with slender flanges is proposed using mixed variatonal method. The current provision for flange local buckling strength for slender flanges is function of web width-thickness ratio. The provision could not consider the effect of moment gradient pattern and flange width-thickness ratio. However, proposed formula could consider flange and web width-thickness ratios and moment gradient pattern. The proposed formula is simple but gives very accurate value for critical buckling stress.