Assessment method for the (remaining) fatigue life of steel bridges with weld imperfections
Looptijd
April 2021 - April 2026Partners
Project manager
Fatigue of steel bridges limits their service life. If a fatigue crack is not detected in time, a crack may grow through the entire structure, which may then collapse. Many existing steel bridges that were built around the years 1960 and 1970 have been found to contain large weld imperfections that are far beyond tolerable according to modern standards. The fatigue performance of weldments with such imperfections is unknown. These bridges may approach their end of fatigue life and can result in an increased demand of resources for Rijkswaterstaat and ProRail. This research project assists in spending these public resources efficiently and effectively, by developing a framework for optimized inspection intervals as well as providing an accurate assessment method for the remaining fatigue life, thereby enabling the extension of the safe use of existing steel bridges.
The project consists of four work packages (WP-s). The improved prediction of the (remaining) fatigue life starts in WP1 through an increased accuracy of the size of the imperfections. Based on this size, WP2 develops prediction models to determine the number of cycles associated with crack initiation (nucleation) and subsequently in WP3 with the growth of physically short cracks. At the end of that growth phase, traditional linear elastic fracture mechanics (LEFM) can be used to predict further crack growth until fracture and the entire process constitutes the prediction of the (remaining) fatigue life. Case studies and dissemination are done in WP4. Out of these WP1 and WP2 are being currently carried out in 果冻传媒, while WP3 and WP4 are with TU-Delft and TNO.
Status of the Development
WP1: Sizing Accuracy of Weld Defects
Time-of-Flight Diffraction (TOFD) is an ultrasonic non-destructive evaluation (NDE) technique in which the size of a defect (e.g., weld imperfection, fatigue crack) can be estimated by evaluating the time-of-flight (ToF) of an ultrasound pulse scattered by the imperfection. WP1 is investigating signal processing techniques, namely 鈥渋maging鈥� and 鈥渄econvolution鈥�, to improve the accuracy of the interpretation of TOFD measurements. Imaging uses ultrasonic signals to reconstruct the inspected geometry and visualize the defect position. Deconvolution is used to sharpen the image, hence increasing the resolution and accuracy of defect positioning.
WP2: Fatigue Crack Nucleation Model for Weld Defects
Traditional approaches for estimating the remaining fatigue life of structures do not take the crack initiation phase into account, of which the crack nucleation stage can be a considerable portion. This approach can be sometimes too conservative. This project aims to develop a modeling approach for predicting crack nucleation from weld defects. Estimating crack nucleation from weld defects is a multiscale problem wherein the shape and size of a defect, the material microstructure in its vicinity, and the fluctuations in the local strain (and stress) field are the governing factors. To address these aspects, a two-scale modeling approach is being developed for predicting crack nucleation from weld defects. Conventional continuum-based finite element modeling is being used in a macroscale model to capture the non-homogenous deformation near the defect, while crystal plasticity finite element modeling is being used in the mesoscale model to predict the crack nucleation in the microstructure.
