Thumbnail Image
Ji, Koochul
Arson, Chloé
Associated Organizations
Supplementary to
In 2017, the overall grade of U.S. infrastructure was D+, which means that most of theinfrastructure was judged structurally deficient. With the current structural ageing rate andmaintenance frequency, almost half of the national bridges will require a major structuralinvestment within the next 15 years. Naturally, the importance of preventive design (exante) and maintenance (ex post) was stressed in many previous studies that aimed to assessreparation techniques. Due to its economical and practical benefits, polymer injection iswidely employed to repair cracks in concrete structures. In this thesis, we investigate themechanisms of mechanical recovery in concrete repaired by epoxy at atomic and metricscales. The first part of the thesis presents Molecular Dynamics (MD) models of concrete/epoxyinterfaces. We focus on High Molecular Weight Methacrylate (HMWM), which is anepoxy resin commonly used in construction. We first propose an MD procedure to buildcross-linked HMWM. The MD model can reproduce the mass density and the glass tran-sition temperature measured during annealing tests reported in the literature. MD pull-out tests on calcite/HMWM and silica/HMWM interfaces show that the tensile strengthof concrete/HMWM interfaces is optimal in dry conditions and at low temperatures. Sil-ica/HMWM interfaces are stronger than the calcite/HMWM interfaces. MD simulationsare conducted at a high strain rate due to computational cost limitations, but Richeton’smodel and Johnson-Cook model are employed to predict the tensile modulus of HMWMand the interfacial strength between HMWM/concrete minerals at a low strain rate. Inorder to investigate whether or not interlocking plays an important role in interface shearstrength, we simulate shear deformation tests with silica/polymer interfaces, in which thesubstrate is either smooth or rough (with and without a notch). On the basis of extensiveparametric studies, we show that the filling ratio of the interface model increases as temperature and injection pressure increase and as the polymer chain length decreases. Interfaceshear strength does not depend significantly on temperature, unless the temperature reaches400K, which is the melting temperature of the polymer. Under same conditions, the shearstrength of an interface with a notch is 1.5 times larger than that of an interface without anotch. In both mode I and in mode II, MD results indicate that the work of separation is85% non-bonded energy, i.e., the resistance to interface failure is mostly attributed to vander Waals forces. The second part of the thesis provides insights into the mechanical behavior of repairedconcrete at the metric scale. We present a numerical modeling approach based on the FiniteElement Method (FEM), in which HMWM joints and cracks repaired by HMWM are rep-resented by cohesive zone elements (CZEs) with bilinear softening, embedded in a meshof volume elements that are assigned a damage-plasticity model. The model is calibratedagainst experimental results obtained on cut and sealed concrete specimens and verifiedagainst data generated from testing reinforced concrete (RC) beams and Pre-Stressed Con-crete (PSC) beams. Experimental data on cut and sealed concrete and on repaired RCbeams was generated by Dr. Stewart’s group at Georgia Tech, and data on non-repairedPSC beams was found in the literature. The RC beams were subjected to a three-pointbending test to produce cracks, and cracks were filled with HMWM by gravity filling. Therepaired RC beams were reloaded until failure. We present a method based on Digital Im-age Correlation (DIC) to identify the zones of high maximum principal strain generatedafter the first loading cycle. We simulate the second loading cycle for several reparationscenarios, where the HMWM filling ratio is varied in the zones of high strain. Simulationresults suggest that HMWM can penetrate cracks of width 0.01 mm and above by gravity.We also find that HMWM reparation increases concrete stiffness and strength if cracks inconcrete members are over 0.1 mm in width, in which case, the load capacity of repaired RCbeams is 30 to 40% higher than that of as-built RC beams. We also simulate pre-stressing, strand release, and four-point loading of PSC girders. We find that the load capacity of aPSC girder damaged by pre-stressing and then repaired would be about 7% higher than thatof the as-built PSC girder. At the same volume fraction of rebar, an RC girder has a loadcapacity that is about half of that of the PSC girder. The addition of HMWM in a zone ofdamage similar to the pre-stressing damage zone would increase the load capacity of theRC girder by 50%, reaching 75% of the load capacity of the as-built PSC girder. Comparedto the as-built girders, pre-damaged girders exhibit a loss of stiffness of 20 to 25% and aloss of load capacity of up to 15%. Results show that reparation allows recovery of bothstiffness and strength. Repaired girders sometimes even exhibit a higher load capacity thanthe as-built PSC girders.
Date Issued
Resource Type
Resource Subtype
Rights Statement
Rights URI