< Back to previous page

Project

Vibration-based prediction of fatigue damage accumulation in metallic structures by use of virtual sensing

Fatigue of metallic structures occurs when a structural component is subjected to cyclic loading.This problem arises for many engineering structures, such as railway bridges, wind turbines, and offshore structures. For existing structures, on site strain measurements allow to verify assumptions made in the design, including the response under dynamic loading, the number of load cycles the structure is subjected to, etc. This may allow for early detection of damage  ccumulation and preventing catastrophic failure. If the number of points critical for fatigue assessment is large, or when these points are not accessible for direct strain measurements, the strains in the critical points can be obtained indirectly from virtual sensing. The strains used in the fatigue assessment are estimated from a limited set of response measurements and a dynamic model of the structure. Although promising results have been obtained in a number of case studies involving data obtained from laboratory and in situ experiments, a number of fundamental challenges need to be overcome before strain estimation can be embedded in a continuous fatigue assessment. Modeling errors and noise on sensor data lead to a high level of estimation uncertainty. This project aims at overcoming these challenges through the development of a robust and innovative methodology for strain measurements using virtual sensing. The methodology will be validated using large-scale experiments on a railway bridge.

Date:1 Oct 2017 →  30 Sep 2020
Keywords:metaalmoeheid
Disciplines:Metallurgical engineering