Title: Unified Shrinkage Model for Concrete from Autogenous Shrinkage Test on Paste with and without Ground-Granulated Blast-Furnace Slag

Author(s): Ya Wei, Will Hansen, Joseph J. Biernacki, and Erik Schlangen

Publication: Materials Journal

Volume: 108

Issue: 1

Appears on pages(s): 13-20

Keywords: aggregate; autogenous shrinkage; ground-granulated blast-furnace slag; hydration products; self-desiccation; shrinkage modeling

DOI: 10.14359/51664211

Date: 1/1/2011

Abstract:
Autogenous shrinkage development was studied as a function of time for cement paste and concrete hydrating under sealed conditions at room temperature. Effects of water-cementitious material ratio (w/cm) (0.35, 0.40, and 0.45), ground-granulated blast-furnace slag (GGBFS) content as a percentage of total cementitious material (0, 30, and 50%) by mass, and aggregate content (40%) by volume on shrinkage development was obtained. Shrinkage measurements started after 10 hours and lasted up to 90 days. Self-desiccation (that is, reduction in pore humidity) was predicted using the HYMOSTRUC model. The effects of w/cm on shrinkage development can be normalized from shrinkage versus pore humidity curves for portland-cement paste. The aggregate effect on autogenous shrinkage was found to follow a Pickett model developed for drying shrinkage. These results suggest that a unified shrinkage model, which combines autogenous and drying shrinkage, exists. The framework for such a model is presented, which incorporates relative humidity (RH), aggregate content, and restraint factor as major variables. GGBFS initially reduces shrinkage as it behaves as a filler, thus increasing the effective w/cm. While the long-term shrinkage is increased, the major factor is most likely a reduction in pore humidity associated with pozzolanic reactions. Due to a higher internal RH in a 0.45 w/cm system, the pozzolanic effect on autogenous shrinkage is more pronounced at later ages.