Creep of synthetic fibre reinforced concrete.

As the most widely used construction material, concrete is constantly in development with continuous advances made in the fields of material science, concrete technology and reinforcement systems. The development of fiber reinforced concrete (FRC) has significantly impacted all three fields. Since the first trials with FRC in the 1960s, the material has undergone significant improvements in both the quality of the materials as well as the understanding of its structural behavior. Nevertheless, despite nearly 50 years of development, research on FRC has been focusing almost exclusively on its fresh state properties and on its short-term characteristics. Additionally, the composite material FRC has often been the sole subject of experimental campaigns, rather than the constituents and FRC together. Finally, the use of FRC is not yet widely incorporated in design codes, but typically, structural design approaches consider only the short-term strength and assume FRC to be a one-phased material. In this thesis, both simplifications are abandoned with the aim to investigate the creep behavior of cracked FRC in a multi-scale and two-phased approach. To that end, a combined methodology consisting of experimental work, numerical simulations and sectional analyses is adopted. 

Firstly, an extensive experimental campaign consisting of nearly 300 tests is completed on two types of polypropylene FRC. The short-term and creep properties are investigated at three different scales: fiber, fiber-matrix interface and the composite scale. The aim of the short-term tests is to characterize the material under various types of loading or with various test parameters. The creep tests investigate the long-term behavior at different load ratios, for time scales up to 9 months. The current study represents the first time such an extended time scale is considered in a multi-scale approach for the creep of FRC. The experiments highlight the importance of the fiber properties and the creep load ratio on the observed behavior across all scales.

Secondly, a numerical model is developed to describe and predict the creep behavior of cracked FRC, based on the short-term and creep properties of its constituents (fiber and fiber-matrix interface). To that end, a two-phased finite element model is constructed where the fibers of a so-called fiber set are considered separately from the concrete matrix. Different fiber sets are generated and a Monte-Carlo analysis is performed to statistically assess the influence of the fiber distribution on the creep behavior. This is the first time such a two-phased numerical approach is adopted to determine and predict the creep behavior of cracked FRC. The simulated creep behavior after 15 years predicts no structural failure and the average fiber stresses amount to only 10 % of the tensile strength. The initial deformation and first-week creep is overestimated by the model.

Thirdly, the structural behavior of a cracked FRC section under sustained flexural loading is analyzed and discussed. A three-stage approach is presented to consecutively describe monotonic, cyclic and time-dependent flexural loading. For the latter, the flexural creep deformations are based on uniaxial creep data from the Model Code and the numerical analyses. Therefore, a more physical basis is provided to determine flexural creep, rather than relying on a phenomenological description with spring-dashpot systems. The proposed calculation method predicts the stress and deformation profile over the height of the cracked section and in time. The predicted data is compared to and validated with experimental measurements. A good agreement is found in the location of the neutral axis, the deformation profile and established inverse analysis methods for monotonic bending. As such, the proposed model can be used to predict the flexural creep deformations and time-dependent CMOD growth based on uniaxial creep data.