Limestone calcined clay cements (LC3) are a promising alternative to traditionally used ordinary Portland cement (OPC). So far, little research has been performed on the early-age creep properties of LC3. This motivates a multi-technique experimental study of three binders: an OPC, a binary blend called limestone Portland cement (LPC) containing, by mass, 70% OPC and 30% limestone, as well as an LC3 containing, by mass, 70% OPC, 15% limestone, and 15% calcined clay. Quasi-isothermal differential calorimetry is used to study the reaction kinetics. Macroscopic mechanical tests comprise uniaxial compressive strength tests, ultrasonic pulse velocity measurements, and hourly performed three-minute creep tests, to characterize the material strength, the elastic stiffness, and creep properties from 1 to 7 days after cement paste production. 870 creep tests are evaluated individually to identify the evolution of the elastic modulus, the creep modulus, and the creep exponent. The existing test evaluation protocols [https://doi.org/f8hgjp, https://doi.org/czmn] are extended to explicitly account for shrinkage strains. LC3 was found to be less creep active than LPC, slightly more creep active than OPC from 1 to 3 days after paste production, and less creep active than OPC from a material age of about 3 days. Macroscopic creep and shrinkage of OPC, LPC, and LC3 are qualitatively related to the microstructural phase evolution of the binders, which is studied with CemGEMS.

The demand for accurate characterization of slag-based CEM II concretes is becoming increasingly important, as the construction sector shifts towards eco-efficient materials. In the present contribution, the basic creep behavior of mature slag-based CEM II concrete as used for Tubbing production at Koralm tunnel in Austria, is identified from classical macroscopic creep tests, and traced back to mixture-invariant characteristics of the hydration products, by means of an adaption of a well-validated multiscale micromechanics model for CEM I concrete. The model is based on an extension of Powers’ hydration model towards consideration of slag, and on an isochoric creep function for the CEM II hydration products. The creep function considers (i) short-to-long-term creep through a piecewisely-defined function comprising a power law for the short-term portion and a logarithmic law for the long-term portion, (ii) temperature dependence of hydrate properties, and (iii) the influence of the internal relative humidity. Downscaling of mechanical properties from the concrete level to the scale of the hydration products evidences a shear creep compliance of CEM II hydration products, which exceeds, by a factor of two, the one known for ordinary Portland cements, CEM I. For the purpose of model validation, the identified shear creep compliance of CEM II hydration products is used for predicting the creep of another slag-based CEM II concrete with a different composition. It is concluded, that slag-based CEM II concretes are especially suitable for applications, where a faster stress relaxation under displacement-controlled conditions is beneficial, e.g. precast segmental tunnel linings.

Graphene oxide has gained substantial attention as a carbon-based nanomaterial in cement-based composites. The present study is focused on the effect of graphene oxide on the early-age creep properties of three cement pastes. The first paste, serving as a reference, is produced by mixing distilled water with an ordinary Portland cement which is free of tricalcium aluminate, at an initial water-to-cement mass ratio amounting to 0.42. The second paste is produced by adding a polycarboxylate superplasticizer to the recipe of the first paste. The dosage of the superplasticizer amounts to 0.09% by weight of cement. The third paste is produced by adding graphene oxide to the recipe of the second paste. The dosage of the graphene oxide amounts to 0.09% by weight of cement. Uniaxial compressive strength tests and ultrasound pulse transmission experiments, evaluated by means of the theory of elastic wave propagation through isotropic media, are performed at material ages of 1, 1.5, 2, and 3 days. Hourly three-minute creep testing is carried out from 1 to 3 days after material production. The execution of the non-aging linear creep experiments and their evaluation within the framework of the linear theory of viscoelasticity follows the developments of Irfan-ul-Hassan et al. [https://doi.org/f8hgjp]. The obtained results show that both the dynamic ultrasonic tests and the quasi-static creep tests deliver virtually the same moduli of elasticity, while neither the superplasticizer nor the combination of superplasticizer and graphene oxide have a significantly influence on to the early-age evolution of the uniaxial compressive strength, the elastic stiffness, and the creep properties.

Concrete exhibits linear creep behavior under uniaxial compression for stresses smaller than some 40% of its strength. Above this level, nonlinear creep occurs. Herein, an analytical model is formulated, explicitly separating the contributions to nonlinear creep from (i) viscoelastic phenomena, and (ii) cracking-induced damage. A re-analysis of a multilevel creep test on a mature concrete [http://doi.org/c795rc] is performed. During the test, the creation of microcracks was monitored by means of the acoustic emission technique. The obtained experimental data are used as input to formulate the analytical model as follows. A validated multiscale model for strength upscaling is used to quantify the hydration degree of the concrete based on its reported strength. This is then used in combination with the reported mix design to predict the linear creep properties of the concrete using another multiscale model valid for nonaging basic creep of cementitious materials under water-saturated conditions. A creep reduction factor is introduced to account for a relative humidity smaller than 100%. The affinity concept [http://doi.org/dzsxwf] is used to estimate nonlinear viscoelastic properties of concrete. Innovatively, damage of concrete is explicitly accounted for through a newly introduced damage factor describing the relation between microcracking to the corresponding decrease in concrete stiffness. Cracking events happening both during quasi-static loading and during sustained loading are shown to have an important influence on the behavior of the specimens. The developed model is able to reproduce also other non-linear creep experiments very accurately up to stress levels beyond the limit of applicability of the affinity concept. This confirms that nonlinear viscoelastic phenomena govern the creep behavior at medium stress levels, while cracking-induced damage dominates the behavior at high stress levels.

Creep and shrinkage are an essential part of the material properties of concrete. To describe the time-dependent behaviour of concrete structures with mathematical models, a huge number of experimental tests is needed to calibrate and improve these models concerning creep and shrinkage of concrete.

Most of the creep and shrinkage tests which are available in the literature were conducted on small sized specimens which are tested under constant environmental conditions. The major goal of a research project conducted by the TU Wien since 2017, is to perform creep and shrinkage tests on large-scale concrete specimens under real environmental conditions. Therefore, large-scale prismatic concrete specimens with cross-sectional areas of up to 1m² were produced and exposed to real environmental conditions. To observe the stress induced creep strains, half of the large-scale specimens were loaded with a post tensioning system. The other half of the large-scale specimens were load-free to measure the load independent eigenstrains (consisting of the shrinkage strains, thermal strains and strains due to cracking). The continuous measurement of the concrete strains is ensured by means of vibrating wire strain gauges.

After more than six years of measuring, the vibrating wire strain gauges showed that they are capable for long-term concrete strain measurements. In the presentation, the usage of vibrating wire strain gauges and the creep and shrinkage measurements of the whole measurement period are presented. Finally, the influence of different environmental conditions (due to different production dates) is discussed.

Thin ultra-high performance concrete (UHPC) overlays offer a protective layer to extend the service life of transportation infrastructures such as bridge decks. The restrained shrinkage in bonded concrete systems can, however, cause stresses that eventually lead to overlay cracking and/or interface debonding. While a transition from normal concrete (NC) to UHPC overlays has helped address several performance issues in bonded concrete composites, questions regarding the impacts of the overlay geometric features (e.g., thickness and slope) and mixture composition (e.g., fiber inclusion and use of recycled cementitious materials) on the development of restrained shrinkage stresses and potential failures remain. These standing questions motivated the current study, in which a set of experimental tests was designed to characterize the overlay shrinkage and determine the composite possible failure modes. The experimental test campaign systematically covered the effects of overlay geometry and mixture characteristics. A set of UHPC-NC specimens were cast with overall dimensions of 180×75×15 cm³ and slopes up to 5%. Integrated distributed fiber optic sensors (DFOS) were employed to monitor the strains inside the bonded UHPC overlays for a period of six months. The collected data provided firsthand information on the magnitude and spatial distribution of restrained shrinkage as well as potential crack formation in the UHPC-NC composites over time. By addressing the potential early-age shrinkage issues, this study outcome is expected to optimize the use of thin UHPC overlays, especially to protect NC substrates.