Time-dependent flexural behaviour of prestressed alkali-activated concrete (AAC) girder with cast-in-situ AAC topping

Abstract

Alkali-Activated Concrete (AAC) is considered as a promising alternative to conventional Portland Cement Concrete (PCC) due to its potential to reduce environmental impacts. However, its application in practical engineering is limited by, among others, insufficient understanding of the long-term structural behaviour of reinforced and prestressed AAC elements. To address this, a series of experiments were conducted on composite girders to investigate the long-term flexural behaviour. The composite girder is formed by a prefabricated prestressed AAC inverted-T girder with cast-in-situ ACC topping concrete. The midspan deflection of two composite girders, subjected to self-weight and additional sustained loading, were measured over a 9-month period. Subsequently, flexural tests under four-point bending configuration were performed at the age of 9 months and the reference age of 28 days. The results showed that the specimens tested at 9 months exhibited reduced initial stiffness, decreased cracking load and larger crack widths in the precast prestressed girder compared to those tested at 28 days. The reduction in stiffness likely stems from decreased elastic modulus and structural cracking. Meanwhile, the lower cracking load arises from prestress losses caused by ongoing (restrained) shrinkage and creep, consistent with AAC material test observations. Larger crack widths observed in the precast girder may result from a degradation of bond between AAC and prestressing strands over time. The distinct failure patterns of the 9-month specimens (anchorage failure for sample subjected to self-weight only and flexural failure for sample exposed to additional sustained load), highlighted the role of creep on bond behaviour between prestressing strands and AAC, particularly as a function of varying stress levels at the level of strands. Finally, analytical models were applied to evaluate the prestress loss and flexural behaviour of the specimens. The effective prestressing force and cracking loads at both testing ages were overestimated when the effects of (partially) restrained deformations between precast and cast-in-situ AAC were neglected. More accurate analytical predictions were achieved when these long-term effects and the level of restraint in the composite girder were considered.