Download PDF

Cement and Concrete Composites

Publication date: 2021-01-01
Volume: 115
Publisher: Elsevier

Author:

Babaei, S
Seetharam, SC ; Dizier, A ; Steenackers, G ; Craeye, B

Keywords:

Faculteit Toegepaste Ingenieurswetenschappen, Ingenieurswetenschappen. Technologie, A1 Tijdschriftartikel, Energie en Materialen in Infrastructuur en Gebouwen (EMIB)KW - Optische metrologie, 3D-ontwerp en Mechanica (Op3Mech), Science & Technology, Technology, Construction & Building Technology, Materials Science, Composites, Materials Science, Hardened cement paste, Drying shrinkage, Poroelasticity, Disjoining pressure, Surface free energy, Multi-mechanism shrinkage, Homogenization, Multiscale, C-S-H, BOTTLE-HYDRATED CEMENT, EFFECTIVE STRESS, CONCRETE, MODEL, SORPTION, WATER, MICROSTRUCTURE, POROMECHANICS, REFINEMENTS, 0904 Chemical Engineering, 0905 Civil Engineering, 1202 Building, Building & Construction, 3302 Building, 4005 Civil engineering, 4016 Materials engineering

Abstract:

A new analytical framework that relies on minimal inputs and combines a number of existing techniques to estimate reversible drying shrinkage strain of OPC-based materials is presented. This includes a multiscale framework for estimating water (de)sorption isotherm (WSI), an analytical homogenization technique to estimate bulk modulus, and a multi-mechanism based drying shrinkage formulation. The minimal inputs needed are the cement composition, microstructural information and mechanical properties of hydrated phases of hardened cement paste. A pore network model that forms the core module of the multiscale WSI provides a quantitative basis for the drying shrinkage formulation. The unique feature of the framework is that only two calibration parameters are involved: (i) a geometric parameter used in the pore network model, and (ii) a constant in the disjoining pressure relationship, which is set to unity mainly due to a lack of knowledge. Importantly, there is no need to calibrate these parameters for every experiment. Results from the framework are compared against shrinkage data from literature that encompass both virgin materials (samples that have never been dried prior to the test) and non-virgin materials. A reasonably good correspondence has been achieved with respect to the non-virgin materials, whereas, the results for the virgin materials are examined mainly to gain qualitative understanding of the role of the microstructure on irreversible deformation and thus to propose a phenomenological model.