Residual life prediction of steel components under high-cycle fatigue based on full-field measurement
Bridges, both steel and concrete, can fail due to ageing. In Germany, about 40% of the trunk road bridges were built between 1965 and 1979. Only a few of these bridges have been designed with ageing at that time. Ageing can be triggered mechanically, chemically, biologically, and under there interaction. Among all, mechanical ageing plays an important role though the whole aging procedure.
Mechanical ageing involves mainly the fatigue, which is basically divided into Low-Cycle-Fatigue (LCF) and High-Cycle-Fatigue (HCF), depending on the number of load cycles at failure. Despite the LCF due to e.g. strong wind load, bridges are mainly affected in the service condition under HCF, whereby the structural component is cracked in macro scale without any prior notice.
The evaluation of the change of state of the component can be done by different methods based on Wöhler line, by which cyclic tests are carried out for one specific construction detail. The relationship between the (nominal) stress difference and number of load cycles represents the fatigue behaviour mathematically through the so-called S-N curve. With an appropriate model for loads and damage accumulation, a life prediction is possible.
However, these methods are not suitable for condition assessment and life prediction of components under HCF on existing structures. Firstly, HCF is a highly localized phenomenon. with the Wöhler test showing a considerable scattering. Secondly, only strains can be measured on site, e.g. with strain gauges. Without knowledge of the material parameters, which can change as a result of fatigue, no stresses can be resulted. Finally, the stress history of a structure cannot be determined. The pre-damage is therefore unknown.
With plane strain measurement, local information can be obtained on the component surface where fatigue usually occurs. ESPI (Electronic Spekle Pattern Interferometry) and DIC (Digital Image Correlation) are available for this purpose. The ESPI system is used here because it is considered to be suitable for detecting material changes due to fatigue due to its high sensitivity to the smallest deformations.
The Virtual Field Method (VFM) is used to solve the inverse problem. This is an inverse method with which the material parameters can be derived with known component geometry, load, constitutive equation and measured strain field. The material model can be linear or non-linear, homogeneous or heterogeneous, and isotropic or anisotropic. The applied virtual (strain) fields are decisive for the solution of the inverse problem. An appropriate selection of the virtual fields can minimize the white-noise in the measurement and reduce the calculation effort significantly.
The aim of the research project is to develop a phenomenological model for steel components to describe the material change or the change of the strain field due to HCF. The model results basically from the measured strain fields up to the first crack. The model should be able to describe the crack initiation phase. The following crack growth can be represented by fracture mechanical methods. The mathematical representation of the phenomenological model should have a global elasticity and maintain the energy balance. The lifetime of a component can be predicted with the model and the surface strain data measured on site.
K. Ritter and K. Thiele. Zur frühen Detektion von Ermüdungsrissen mithilfe der Speckle-Interferometrie. Tagungsband des 21. DASt-Forschungskolloquium, 2018.
K. Ritter and K. Thiele. Monitoring Micro-damage Evolution in Structural Steel S355 using Speckle Interferometry. Proceedings of the 7th International Conference on Fracture Fatigue and Wear, Ghent, Springer Verlag, 2018.