PSI - Issue 64

Antonio Bossio et al. / Procedia Structural Integrity 64 (2024) 56–64 Author name / Structural Integrity Procedia 00 (2019) 000–000

57

2

1. Introduction This research builds upon existing studies that used the Finite Element Method (FEM) to analyze and model corrosion in concrete (Lignola et al., 2009, Bossio et al. 2017, 2018). These FEM analyses explored the possibility of applying results from simple "single bar" simulations to more complex Reinforced Concrete (RC) elements where internal pressure is not uniform (Lignola et al., 2014). The authors propose a more detailed model to assess the non-linear stress distribution within concrete as reinforcement bars corrode. They introduce two separate mechanical analytical models with multiple layers. The first model simulates the impact of corrosion until a crack begins to form, while the second model tracks crack growth until it reaches the concrete surface. Both models enable investigation of how concrete and steel key properties affect the cracking process in RC members, while also considering the reduction in bar diameter caused by corrosion. A significant decrease in bar diameter (down to the level of the rib height, h rib ) can lead to bar detachment from the concrete. The expansion of corrosion products has a two-fold effect on the concrete-bar bond. Initially, it increases the bond due to higher internal pressure, but subsequently weakens it as cracks develop in the concrete. This study aims to evaluate how the corrosion process impacts bond development in RC structures. By solving a non-linear system with four layers, the authors calculate the specific level of corrosion penetration ( x ) that triggers crack initiation and propagation within the concrete. Theoretical plots will then be used to illustrate the influence of corrosion on bond development. 2. Background 2.1. Modeling parameters Concrete is assumed to be elastic when compressed (due to its reduced stress level) but fully nonlinear in tension. A simplified bilinear model is used to represent concrete in tension, with an initial elastic stage up to peak tensile strain ε ct and then a decrease in stress until it reaches zero at an ultimate strain ε u . The model considers factors like crack opening and the size of the largest aggregate in the concrete. It also follows Model Code 2010 (MC10) and considers different bar diameters (Φ10, simulating stirrups, Φ16 and Φ20 simulating longitudinal reinforcement) and concrete cover thicknesses. Finally, the effect of time (creep) and the type of aggregate on the concrete's stiffness are included to assess how they affect corrosion in reinforced concrete members. 2.2. Previous Finite Element Analysis Lignola et al. (2009) used FEM analysis to see how well results from lab tests on single rebar specimens compare to real concrete members with multiple rebars. The simulations, for both single bars and real members, were able to check how different factors like concrete strength, rebar size and spacing, and concrete cover thickness affect cracking in the concrete. They started by modelling a single bar surrounded by a cylinder of concrete (like a "single bar" lab test). Then, they looked at real reinforced concrete elements and proposed some basic rules to connect the results from the simple model to the more complex real-world case. By running different simulations, they found that the simple model with a single bar only gives slightly inaccurate results depending on concrete cover thickness in the model. This thickness needs to be the minimum of two values: the actual cover thickness and half the spacing between the rebars. With this adjustment, the underestimation from the model becomes small, around 10%, which is similar to the typical uncertainty when measuring the concrete's tensile strength. 2.3. Cracking initiation A more complex model to understand how stress changes inside the concrete cover as the rebar gets thinner due to oxide was proposed (see Figure 1a). Oxide creates a layer around the rebar, and this model relates the amount of oxide