A mechanistic model for the time-dependent autogenous shrinkage of high performance concrete. (20th September 2020)

• Record Type: Journal Article
• Authors: Huang, Le; Chen, Zhijian; Ye, Hailong
• Is Part Of: Construction & building materials. Volume 255(2020)
• Publication Date: 2020-09-20
• Subjects: Time-dependent model -- Autogenous shrinkage -- High performance concrete -- Capillary tension -- Silica fume; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2020.119335
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: The time-dependent autogenous shrinkage behavior of HPC is modelled. The effect of silica fume on the development of autogenous shrinkage is considered. The development of elastic modulus of concrete over time is considered. Abstract: An accurate prediction of time-dependent autogenous shrinkage behaviors of concrete, especially for high performance concrete (HPC) at an early age, is of great significance to assess and control the cracking risks of restrained structural elements. In this study, based on the capillary tension theory, a mechanistic model for evaluating the time-dependent autogenous shrinkage behaviors of high performance concrete is proposed. A total of 416 data points including the concrete composition, curing condition, age of concrete, water-to-cement (binder) ratio, internal relative humidity, elastic modulus, and measured autogenous shrinkage are selectively collected for the model establishment. The effects of silica fume on the development of autogenous shrinkage are also considered. Upon the sound physical basis, the model requires only a few parameter inputs related to the mixture proportion and physicochemical properties of constituents. The reasonable agreements between the analytical predictions and independent experimental results, as well as common used formulas from different codes (i.e., ACI 209, Eurocode 2, and Model Code 2010), demonstrate that the time-dependent evolution of autogenous shrinkage of HPC can be reasonably predicted by the model proposed in this study.