Steel Structures: Considerations to Reduce Failures Due to Instability

This chapter presents case histories related to failures during construction and service. Eleven stories of a 28-story structure in Los Angeles collapsed during erection circa 1985 when steel beams and columns, yet to be erected, were temporarily placed on erected metal deck and steel framing. Several failures of open wrought iron bridges with “pony trusses” occurred in Western Europe and Russia in the late nineteenth century. Insufficient lateral rigidity of the upper chord in compression led to the failures. There also was a lack of understanding of the flexural rigidities of the top chord and vertical members. One of three cantilevered steel tapered girders, supporting a two-story wood framed house overhanging the beach adjacent to the Pacific Ocean, failed during a major storm. This example illustrates how corrosion can not only significantly reduce the strength of members but also can increase instability primarily owing to the reduction in thicknesses.

In researching the nearly 300-year history on the testing and development on buckling and instability, the author found it intriguing to learn how long theories have taken to be established. The well-known Euler Theory on the elastic behavior of axial loaded members took about 100 years to be validated and accepted. This chapter provides the historical overview of the subject of buckling and instability. It includes representative formulas to illustrate the development of buckling theory. A relatively simplified method, considering an eccentrically loaded column, developing a formula known as the Secant Formula was proposed in 1858. It took into account initial out-of-straightness of columns. Bracing of compression flanges of beams and girders may initially be looked at simply as a column. However, whereas column buckling is a function of flexure with beam buckling, both flexure and torsion occur, which affect bracing requirements.

The requirements for external and internal forces in general are well-defined in codes and design manuals. The significant development of computer analysis software programs, carried out over the last six decades, has appreciably improved the ability to analyze complex structures and assess second-order effects. Postprocessing software programs for elemental design have further supplanted many of the calculations previously carried out manually. In many instances, steel members and their connections are expected to undergo limited to significant yielding during a seismic event. Furthermore, members and their connections are subjected to cyclic forces with both elastic and plastic behavior occurring. Stability considerations include the following: global stability of the structure, global stability of individual members, and local buckling (e.g., member flanges and web). Regarding steel moment frames, which are expected to be subjected to significant yielding, the difficulties in controlling buckling mechanisms can be highly significant.

The steel making process can have significant influences on steel properties across the section of the steel member. During the rolling process, plastic deformation takes place such that the material is work-hardened. The steel-making process considers the chemistry, rolling practice, and thermal to control the desired steel properties. The stress-strain properties of steel can have an appreciable effect on buckling performance. It has been well established that welding, such as for plate girders and box sections, can cause high residual stresses at or near the welds. Connection details including beam to column connections, splice details, and braced frame details, among others, and can cause unintended eccentricities that reduce buckling and lateral torsional buckling strength of members. Often, beam copes are not fully identified in the design process but established during the shop drawing process.

This chapter presents measures to minimize risk of failures during service. These include recognizing eccentric conditions, three-dimensional effects, and out-of-plane distortions. In addition, the chapter discusses changes owing to temperature and shrinkage, bracing for strength and stiffness, redundancy, independent check/peer review, and further research and development. The chapter also presents measures to minimize risk of instability during construction. These include structural or civil engineers’ responsibilities, processes to minimize failures during construction, and liability issues. Furthermore, it is highly important that the education of structural engineers covers in depth the historical development of the extensive methods of evaluating instability so that they have a thorough understanding of the subject and not merely adhere to prescriptive code requirements.