Concrete fatigue modeling and characterization across scales: lattice-discrete and microplane models rooted in a unified dissipative hypothesis

RWTH Publications (RWTH Aachen)

Abstract

The characterization and prediction of fatigue behavior in concrete structures have become increasingly critical for infrastructure design and lifecycle management, driven by aging structures, sustainability imperatives, eco-efficient binder deployment, and emerging applications such as hybrid towers for wind turbines. While extensive experimental datasets exist for conventional concrete under idealized cyclic loading, fundamental knowledge gaps persist regarding multiaxial stress states, load sequence effects, environmental interactions, and the fatigue response of novel binders. Bridging these gaps necessitates physics-based models capturing degradation mechanisms from first principles. This dissertation presents a multiscale computational framework for concrete fatigue in which a thermodynamically consistent constitutive framework, formulated at the microstructural level of concrete, is systematically calibrated and validated against experimental data across multiple length scales and loading configurations. The theoretical foundation is based on an experimentally rooted hypothesis that fatigue-driven degradation at the aggregate–cement paste interface is governed by cumulative inelastic strain evolution. This hypothesis is formalized within a thermodynamically based framework and implemented as constitutive relations in two complementary paradigms: (i) a discrete formulation within the Lattice Discrete Particle Model (LDPM) package MARS, enabling explicit representation of mesoscale heterogeneity and discrete crack evolution; and (ii) the continuum microplane model MS1, implemented in the finite element code ATENA for structural-scale analysis. Systematic validations are conducted across multiple scales: material point simulations, uniaxial compressive fatigue tests on cylindrical and prismatic specimens, punch-through shear tests imposing multiaxial stress states, and fatigue tests on prestressed concrete beams. An energy-based analysis elucidates load sequence effects, which may yield unsafe predictions when employing the conventional Palmgren–Miner damage accumulation rule. A principal finding is that cumulative damage-induced dissipation up to failure exhibits remarkable invariance across different stress amplitudes and fatigue lifetimes, scaling with the fatigue process zone size. This indicates a characteristic volumetric energy dissipation capacity for a given stress state that is exhausted during fatigue, largely independent of the applied loading sequence. This research advances computational modeling of concrete fatigue by establishing a physically motivated, thermodynamically rigorous framework linking microstructural degradation mechanisms to macroscopic structural response. By integrating first-principles-based constitutive formulations within established numerical simulation platforms, the proposed methodology yields fundamental insights into concrete fatigue mechanisms and provides a rational foundation for reliable and efficient design strategies for fatigue-loaded concrete infrastructure.

Authors 1

1. Mario Aguilar corresponding Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen