Estimating viscoelastic compliance of desiccating cementitious materials using drying prism tests

Highlights

• Determining relative humidity and degree of reaction of desiccating cementitious materials via an integrated method

• Extracting moisture diffusion coefficient of cement mortar using the mass loss data of the prism specimens

• Estimating uniaxial viscoelastic compliance of cement mortar using the shrinkage strain data of the prism specimens

Abstract

An integrated modeling approach to determining internal relative humidity and degree of reaction of desiccating cementitious materials has been developed. By simultaneously accounting for both self-desiccation and external drying, the model is capable of predicting the evolution of material properties of desiccating cementitious materials. Using the mass loss data of the prism specimens, the moisture diffusion coefficient of a nonlinear model for cement mortar was determined. Additionally, a previous poroviscoelastic shrinkage model was modified and extended. Through comparison of predicted and measured axial strain of the prism specimens, the viscoelastic compliance of cement mortar was extracted. The present work revealed that it is possible to use a simple drying prism test to fit the moisture diffusion coefficient and creep compliance of cementitious materials, though the comparison of the extracted viscoelastic compliance with measured compliance was only reasonable if the outer layer of the drying mortar was presumed to have microcracking.

Introduction

Concrete is one of the most versatile construction materials and has many advantages including the potential for high compressive strength. However, concrete structures still fail because of degradation from excess shrinkage, exposure to harsh environment, reactions with common elements, etc. Often, these factors cause far more damage than mechanical loads. A primary source of concrete degradation is due to shrinkage [[1], [2], [3]]. If concrete is restrained, shrinkage induces tensile stress and can eventually cause microcracks that coalesce into macrocracks, which might jeopardize the structural integrity.

Cementitious materials may generally be classified as partially saturated porous media. Upon loss of pore water, concrete shrinks because of pore fluid pressure changes (desiccation shrinkage). Although the loss of pore water often comes from two sources—moisture transport from the inner body of concrete to its surfaces that are exposed to a drier environment (resulting in drying shrinkage) and internal water consumption due to reaction of cementitious materials (resulting in autogenous shrinkage), the shrinkage due to water migrating to open boundaries and the shrinkage caused by self-desiccation only differ in how the pore water is removed, whereas the shrinkage deformation mechanisms are identical (driving pore pressure reductions for both may be approximated by the well-known Kelvin-Laplace equation). Therefore, the concrete volume changes due to self-desiccation and external drying should be modeled in a coupled manner via poromechanics rather than separated. A challenge in modeling volume change of cementitious materials is that concrete exhibits aging, viscoelastic behavior. The effects of aging and viscoelasticity on the deformation are particularly significant for cementitious materials at early ages and may be exacerbated for materials containing different additives, such as fly ash [4] and asphalt coated aggregate [5]. Viscoelastic deformation relaxes a significant amount of stresses in cementitious materials, which often results in delay in major cracking occurrence [[5], [6], [7], [8]].

Grasley and Leung [9] showed that the deformation of cementitious materials due to desiccation can be modeled via poroviscoelasticity by considering aging due to solidification of new hydration products in combination with inherent aging of the cement paste gel. However, this work was restricted to one dimensional and did not couple hydration, self-desiccation, or drying models to the shrinkage model. The objective of this paper is to modify and extend this poroviscoelastic approach by integrating a moisture diffusion model to calculate water loss due to both self-desiccation and external drying. The other contribution of this work is to demonstrate that the developed model can yield satisfactory fitting results for cementitious materials transport properties (e.g., moisture diffusion coefficient) and viscoelastic properties (e.g., viscoelastic compliance) simply based on the drying prism test (ASTM C157).