Effect of crack shape and size on estimating the fracture strength and crack growth fatigue life of bridge cable steel wires

Abstract

Macrofracture entails the creation of free surface. The process can be enhanced or impeded depending on the ways in which the material microstructures are designed to react against the operational conditions. Undesirable chemical reactions can be the main source of strength degradation for load-supporting large structural members. Assessment of the remaining strength of damaged cable wires for cable-stayed or suspension bridges can rely on a knowledge of strength and/or fatigue life depending on the prevailing stress amplitude and whether the cables are stayed or not. Although large-scale computer schemes are available for making detailed failure analyses, they are not conducive to retrieving technical information within a relatively short time. Computer simulation has not advanced to the stage where experimental validation could be spared. To this end, effective analytical/experimental methodologies are much in need of development. The fracture mechanics approach appears to have gained ground in recent times, notwithstanding of its extension to include material microstructure effects. The present work was motivated by the successful completion of the Wuhu Changjiang River Bridge (WCRB) where the fatigue life of the cable and anchorage was a concern as the cable-stayed double-deck bridge can carry heavy loads for both highway and railway transportations. Use was made of the available information on the fatigue of bridge cables consisting of strands of steel wires to develop a fatigue crack growth rate model for examining possible micro-/macro-crack transition that is conducive to eutectoid and hypereutectoid steels with pearlite microstructures that has been drawn to possess very high strength. The static strength range under study for bridge wires is about 1770 MPa. The current state of the art relies on testing the cables with limited studies on the steel wires based on the static strength evaluation. Nevertheless, it is not possible to elude the fact that the fatigue lives of the steel wires have to be qualified by the manufacturers. After all, the cable fatigue lives do depend on those of wires in the bundle. The specification for the WCRB is two million cycles for steel wire while allowing only 2% of the wires to fail for a fatigue life of two million cycles for the cable. The prediction of wire bundle strength from a single wire would be analogous to that for the fibre bundle in composite studies that has been subsided in time. Following a brief account of the current approach for determining the fracture strength of high strength bridge cable steel wires under tension and/or bending, the predicted results are scrutinised against the different shapes and sizes of the crack profiles invoked in the analysis. The assumed crack profiles are also tested by using a fatigue crack growth model for a high strength steel with the same mechanical properties such that the model-dependent fatigue crack growth rate parameters can be determined. Because of the large difference of the cable and wire size, the rules of geometric similarities must be restrained from application to strength evaluation. Bodies with large surface to volume ratios will invariably contain more surface defects in contrast to those with small surface to volume ratios where failure will initiate from the interior rather than the surface. This is precisely the difference between the physical mechanisms of failure in static tension and fatigue. Keep in mind that the surface to volume ratio of the thin wire used for the WCRB differs by an order of magnitude and more for the cable by assuming an equivalent cross-sectional area of the cable to the wire bundle consisting of two cases, one with 73 and other with 85 steel wires of 7 mm in diameter. Both lens shaped and straight line crack front geometries were considered in the application of the fracture mechanics approach. Results for six cases in total, three for tension and three for bending, are obtained. The ranking based on the static fracture strength agreed with the predictions of the fatigue cracking model despite the fact that actual failure mechanisms in the two situations are known to be different as mentioned. The fatigue crack growth rate model, however, has the potential to address the transitional behaviour of microcracking and macrocracking; whether the conditions corresponding to strength measurement coincides with the triggering of a dominant macrocrack with or without the influence of microcracking has been a debatable issue in fracture mechanics known as `plasticity effects’. Aside from using the dual scale model to study the transitional behaviour of micro- and macro-cracking, an equivalent bridge cable with homogeneously distributed wire may be developed to circumvent the irreconcilable difference in size effects of the wire and the cable.