Simplified Constitutive Model for Fatigue Behavior of Concrete in Compression

Abstract

In the literature, three basic assumptions are used to modify monotonic constitutive models in order to simplify fatigue analysis of concrete. First, the fatigue hysteresis loop at failure is assumed to intersect the monotonic stress–strain envelope. Second, it is assumed that the peak stress of a fatigue-damaged concrete element intersects the monotonic stress–strain envelope. Third, the centerlines of fatigue hysteresis loops are assumed to converge at a common point. Although the modifications supposedly lead to improved predictions, experimental verifications of these assumptions are currently insufficient to justify their implementation in the fatigue analysis of complex and large concrete structures where considerations of safety and cost-effectiveness are expedient. From experimental verifications conducted to ascertain the conservative level of these assumptions, it was found that the first and second assumptions seem reasonable, while the third assumption was inaccurate and thus in need of improvement. As such, a new convergence point is proposed. The constitutive models for high-strength and normal-strength concrete in compression were also modified as functions of the irreversible strain and residual strength. Further, a model was proposed for the irreversible strain accumulation, and its corroboration with experimental results showed good agreement.