Automating Crack Inclusion in Finite Element Modelling Using Data from Structural Health Monitoring - Structural Health Monitoring Refinement on Assessing the Structural Behaviour of Reinforced Concrete

Abstract

Inadequate monitoring of structural damage in reinforced concrete infrastructure is one of the most common causes of structural deterioration and at worst, failure, which entail unnecessary risks and costs, both financial and environmental. Furthermore, Finite Element (FE) tools of today use generalised methods to predict, rather than describe, the formation of cracks when evaluating reinforced concrete structures. In an attempt to circumscribe these problems, this thesis aimed to refine the structural assessment of reinforced concrete structures by automating crack in clusion, using data from Distributed Optical Fibre Sensors (DOFS), in FE Analysis (FEA). DOFS, installed inside reference test beams, were utilized to measure strain data during laboratory tests, which was then processed in a series of scripts and incorporated in an FE model. The scripts were written in such a way that the process of locating cracked elements and assigning these their corresponding weakened properties, was automated. It was found that including cracks, translated from DOFS strain data, in FEA gave a far better numerical estimation of crack widths compared to a traditional model without crack inclusion. Moreover, the crack inclusion itself is highly functional and applicable, and conducting further laboratory test with DOFS can further improve the method. Furthermore, up to the load where strain data from DOFS was gathered, an improved representation of the structural behavior could be noted when comparing the FE model with crack inclusion to a traditional FE model without crack inclusion.