Finite-element study of the diagonal-tension failure in reinforced concrete beams

Abstract

In this work, we aim to tackle one of the most devastating failure modes in reinforced concrete (RC) structures: the diagonal-tension failure. In order to study this phenomenon numerically, a model capable of dealing with both static and dynamic crack propagation as well as the natural transition of these two regimes is necessary. We chose a discrete cohesive model for concrete fracture, an interface bond-slip model for the deterioration between concrete and steel rebar, both combined with an insertion algorithm. The static process is solved by a dynamic relaxation (DR) method together with a modified technique to enhance the convergence rate. The same DR method is used to detect a dynamic process and switch to a dynamic calculation. The methodology is applied to model the experimental results of Carmona et al. (Engineering Fracture Mechanics 74:2788-2809, where the recognition of the transition to a dynamic fracture in a presumably static calculation is essential to reproduce the diagonal-tension failure observed.

FEMME: A multi-scale Finite Element Microstructure MEshfree fracture model for quasi-brittle materials with complex microstructures

“…The synthetic microstructure created has a porosity of 1%, which is formed by pores of an average size of 1 mm with a standard deviation of 0.2 mm. Comparison is made figure 19b, in terms of the predicted fracture load (data for the range of strength experimentally measured by Arrera and Ingraffea [27] are shown) and the number of elements, with other methodologies that either need to predefine a fine mesh along the fracture path, such as cohesive faces (DCFF) [2], cohesive elements [3] or strong discontinuities (SDA) [28], and those that can deal with a mesh without a predefined fracture path, such as XFEM and EFEM [5]. The FEMME model is shown to reach a good solution using a coarse mesh, demonstrating the potential for this methodology to simulate fracture using the coarse mesh in structural elements without requiring a predefined fracture path.…”

“…These models simplify the material as a continuum; the appearance of the crack within an element is represented either by a change in its mechanical properties, or the crack is represented at the interface of two elements by mesh splitting and inserting equivalent forces [2]. Such techniques include: cohesive element models [3], which insert homogeneous damage inside a FE; the Strong Discontinuity Approaches, which are evolutions of the cohesive elements method that inserts an oriented crack inside a FE; or the Cohesive Faces model [4], which inserts the crack at the interface of two elements.…”

Finite element modeling of debonding failures in FRP-strengthened RC beams: A dynamic approach

“…In modeling structures made of brittle materials such as concrete and masonry, unstable structural responses involving ''snap-back'' due to crack propagation and/or strain-softening usually exist and may induce local dynamic processes in the middle of a overall static process under certain conditions [51], making it difficult for common static solution strategies such as the displacement-control Newton-Raphson method to obtain a converged solution [45,46,52]. The arc-length method with its different forms has been repeatedly reported to have difficulties in finding a converged solution although the method was coined for obtaining structural responses associated with either snap-back or snap-through phenomena [53][54][55][56][57][58].…”

“…multiple cracks in a concrete member) are activated and interact with each other [61,62]. Therefore, solving for the static response of a structure with strong strain localization phenomena, such as concrete cracking, remains a challenge [45,46,52].…”

“…Tavarez [63] modeled the flexural behavior of concrete beams reinforced with composite grids using the explicit FE method; Yu and Ruiz [46] applied the modified dynamic relaxation (DR) method [64] to model crack propagation in RC beams; Coronado and Lopez [20] employed the explicit FE method in ABAQUS to model FRP-strengthened RC beams. More recently, the explicit FE method was used by Dhanasekar and Haider [65] to model lightly reinforced masonry shear walls and by Zheng et al [66] to model the punching shear failure of bridge decks; the modified DR method was used by Yu et al [52] to simulate both static (stable) and dynamic (unstable) crack prorogations. All of these studies have been successful to different extents, and some of them have obtained predictions in close agreement with test results.…”