Shear Assessment of Critical Concrete Members in Existing Infrastructure

The retrofit or replacement of bridges, transportation networks, and critical civil buildings is a major problem in many countries as a large stock of structures built during high economic periods are rapidly approaching the end of their service life. As a result, in the coming decade a large number of bridges will enter a critical phase exceeding 50 years of service, typically linked to significant deterioration. At the same time, traffic volumes are continuously increasing in intensity and weight of vehicles and goods, posing additional challenges for the safety of the aging infrastructure. When the shear capacity of such structures is assessed, this is typically done based on models and provisions developed primarily for the design of new structures. However, existing concrete structures often have a reserve of shear strength, which is not captured by shear strength methods in design codes. On the scale of large infrastructure, such built-in conservatism can require very disruptive and costly large-scale interventions. In this context, and because resources for strengthening are limited, it has become increasingly more important to enhance the structural assessment methodologies used to identify, prioritize and conduct cost-effective interventions.

Part I of this thesis focusses on the shear assessment of concrete deep pier cap beams in the substructure of bridges. Deep beams are typically heavily loaded members that develop wide diagonal cracks, whose exact geometry has a substantial impact on the shear strength. To reflect these characteristics, Part I presents a framework for the crack-based assessment of lightly reinforced concrete deep beams that uses the observed crack geometry as an input in the Two-Parameter Kinematic Theory (2PKT) to provide member- and crack-specific assessments of structural response. The results indicate that the crack-based assessment framework is able to capture the difference in shear response of nominally identical members that exhibit different crack geometries, and is capable of predicting the stress distribution along the critical cracks.

The assessment framework provides different levels of approximation (LoAs) with progressively increasing accuracy and modelling effort. The first level of approximation (LoA I) involves using the 2PKT in a predictive manner for situations where the crack shapes are unknown, LoA II improves upon LoA I by incorporating more detailed modelling of the strain in the transverse reinforcement at the location of the critical cracks to capture potential brittle ruptures of the stirrups, and LoA III includes the details of LoA II but also uses the measured crack geometry as an input in the assessment to perform crack-specific structural assessments. Furthermore, LoA III is used to study the size effect in shear for deep beams, and to quantify the influence of aggregate interlock on the size effect in shear critical deep beams.

While existing approaches can be used to predict crack information for comparisons with on-site observations, there remains a need to establish methods that use crack measurements as a direct input to assess residual structural capacity. More specifically, three important questions are addressed: 1) which critical crack displacements should be measured?; 2) where should the critical crack displacements be measured?; and 3) what is the residual shear capacity of the member, given the measured critical crack displacements? To answer these questions, detailed data from large-scale experiments is examined and interpreted with the crack-based 2PKT. The results show the importance of vertical crack displacements to determine the residual capacity of deep beams. The crack-based assessment is used to determine the residual capacity of large-scale deep beam tests with variable properties and governing load-bearing mechanisms.
Part II of this thesis focusses on the shear assessment of prestressed concrete slender beams in the superstructure of bridges. In the past, such members were typically designed with post-tensioning tendons to balance permanent loads, to maintain the girders uncracked under service loads, and to ultimately enhance the durability and service life of concrete bridges. The structural assessment of prestressed concrete girders poses serious difficulties related to the increase of loads over the past decades, as well as outdated designs that do not satisfy strength and detailing requirements of modern code provisions. Particularly challenging is the shear strength assessment, which nowadays is carried out according to modern shear provisions based on mechanical models that distinguish between the uncracked and cracked state, and require a certain minimum amount of transverse reinforcement. However, the common practice in the past was to limit the shear demand to the principal tensile stresses in the uncracked state, and therefore only light shear reinforcement is typically found in existing prestressed girders.

To address this challenge, Part II presents an experimental program that consisted of three full-scale monotonic tests of prestressed concrete slender beams extracted from a bridge in Belgium in 2023. The prestressed girders were 30 m-long, 1.90 m-deep, and built with multiple bonded post-tensioned parabolic tendons. The main test variable is the position of the applied point load. Given the complexity of such members, a second variable is studied in the experimental program – the effect of the tendons layout on the load-bearing mechanisms.

The experimental results are analysed, compared, and contrasted to investigate the effect of the test variables on the shear resistance. Global and local deformation measurements, including crack patterns and deformed shapes, at different load stages are reported and examined. The shear strength of the specimens is compared to the predictions of four well-established design codes to identify gaps of knowledge in the context of structural assessment of existing lightly reinforced prestressed concrete members. The behaviour observed during the tests showed that the beams had an important reserve capacity that is not captured in a rational manner by shear models of modern codes. The modelling assumptions are discussed together with the experimental observations, and an extended shear strength model is proposed. The extended model, which builds on the existing shear provisions of the Canadian code for concrete structures, is validated against the three full-scale tests and other large-scale tests from the literature.