Viscoelasticity-induced fracture behavior of rock-concrete interface after sustaining creep process

https://doi.org/10.1016/j.cemconcomp.2022.104901Get rights and content

Abstract

This paper explored the viscoelasticity-induced fracture behavior of rock-concrete interface after sustaining the creep process. Creep tests were firstly conducted on the composite rock-concrete specimens under three-point bending loading at the load levels of 50% and 75% of the maximum load. After 90-day sustained loading, composite specimens were unloaded and then reloaded under quasi-static loading conditions, and the digital image correlation technique was employed to quantify the crack profiles at the rock-concrete interface. The results indicate that the pre-set crack at the rock-concrete interface would not initiate further during the creep tests if the sustained load was below the initial cracking load. After sustaining a creep process, the initial cracking load increased obviously but the elastic strain energy near the pre-crack tip remained unchanged. The creep deformations during the creep tests made the characteristic lengths of the rock-concrete interfaces decreased largely, leading to the obvious enhancement of the interfacial brittleness. As a result, the interfacial nonlinear fracture behavior from the initial cracking state to the unstable fracture state would become weaker. In addition, the fully formed fracture process zone length and the corresponding crack length increased with the increase of the sustained load level, indicating that the boundary effect of the rock-concrete interface would decreased.

Introduction

For a concrete gravity dam constructed on rock foundation, the interface between rock and concrete is normally the weakest link in terms of the load bearing capacity due to the inappropriate bonding between them. Initial defects easily form at the rock-concrete interface during construction process and become a potential threat to the safety, stability and serviceability of the dam [1,2]. In practical engineering, the concrete gravity dam operates under long-term loading caused by the water pressures on the upstream face. In particular, the initial defects at the rock-concrete interface are subjected to the sustained hydraulic fracturing effect [3,4]. In this case, the mechanical features and material properties at the rock-concrete interface keep changing over the loading period [5], which may induce the further propagations of the initial defects and even lead to final failures of the structures [6]. Therefore, the effect of sustained loading should be considered when detecting and assessing the safety, stability and serviceability of a concrete gravity dam during its service life.

According to the existing researches, the time-dependent mechanical behaviors of structures with initial cracks are closely related to the sustained loading levels [7]. In general, when the sustained loading level is low, only the viscoelastic properties of bulk materials govern the mechanical behaviors of the structures, leading to a linear creep [8]. By contrast, when the sustained loading level is high, irreversible damage accumulates near the crack tip during the loading period [9,10]. Moreover, the coupling effect of the damage and viscoelasticity may provoke the continuous crack propagation, leading to the divergent deformation, which is called the nonlinear creep [11]. It is worthwhile to point out that, during the most service life period of the dam, the normal water level cannot drive the initial defects at the interface to propagate [12]. Even so, the initial crack still has potential risk of unstable propagation when a dam suffers from the extremely high load in the subsequent service period. To ensure the normal operation of the concrete gravity dam in the whole service life, it is meaningful to explore the viscoelasticity-induced fracture behavior of the rock-concrete interface.

Similar to concrete and other cement-based materials, the rock-concrete interface in a concrete gravity dam can be also regarded as a typical quasi-brittle material [13,14]. After the initiation of a crack, there exists a fracture process zone (FPZ) ahead of the pre-crack tip to reflect the strain softening property of the interface. Due to the cohesive effect within the FPZ, a stable crack propagation process occurs before the ultimate state is reached [15,16], and the fracture behavior from the crack initiation to the ultimate state exhibits obvious nonlinearity. The nonlinear fracture behavior contributes to a larger deformation before the ultimate status, which is beneficial to analyze the crack stability and forecast the structural risk from the view of a structural monitoring [17]. When considering the viscoelastic characteristics of bilateral materials, creep deformations accumulate during the sustained loading period, and the pore spaces around the crack tip increase largely due to the development of new microcracks and/or material extensions. As a result, the deformability at the rock-concrete interface becomes weaker after sustaining a creep process, and nonlinear fracture property becomes lower in the subsequent loading process. In addition, due to the accumulation of the creep deformations, the total structural deformation is different from that under quasi-static loading. Furthermore, the total structural deformation measured in practical engineering may be inapplicable to analyze the fracture property of the rock-concrete interface based on the linear elastic fracture mechanics.

During the interfacial crack propagation process, the FPZ length is also a significant parameter to reflect the nonlinear fracture property of the rock-concrete interface. Many investigations have been conducted to analyze the FPZ evolution at the rock-concrete interface, either experimentally [13,18] or numerically [19,20]. These results indicated that the FPZ length experienced three obvious stages, i.e. ascending stage, plateau stage, and descending stage. In addition, the fully formed FPZ length increased with the increase of the specimen height, and decreased with the increase of the pre-crack length [20]. The afore-mentioned conclusions about the FPZ evolution at the rock-concrete interface were obtained under quasi-static loading conditions. Regarding the effect of creep deformations, some investigations have been conducted on the FPZ evolution in concrete after undergoing sustained loading. For example, Saliba et al. [21] employed the acoustic emission (AE) technique to detect the FPZ width in the concrete after the four-month sustained loading at 85% of the peak load. Based on the amplitude distribution, the FPZ widths in the concrete specimens showed an obvious decreasing tendency in the post-peak stage compared with the control specimens. Omar et al. [22] revealed that the FPZ length of concrete significantly decreased after the creep tests by introducing the size effect model to analyze the characteristic length. Dong et al. [23] investigated the FPZ length of concrete in different loading stages after the creep tests. It was found that the FPZ length became larger after undergoing sustained loading. Due to the similar fracture behaviors to the concrete materials, the creep deformation of the rock-concrete interface would have a significant influence on the interfacial FPZ evolution. Therefore, to further comprehend the influence of the creep deformation, it is necessary to quantify the crack profiles and then derive the FPZ evolution formulations at the rock-concrete interface after experiencing the creep process.

In line with this, the objective of this study was to investigate the effect of the preceding creep deformation on the fracture behavior of the rock-concrete interface. Creep tests were conducted on the composite rock-concrete specimens with two interface roughness degrees, i.e. 4 × 4 and 7 × 7 profiles, under two sustained loading levels, i.e. 50% and 75% of the maximum load. After the 90-day sustained loading, the creep specimens were unloaded from the steel frames and then reloaded under the quasi-static loading until failure. During the reloading process, the DIC technique was employed to quantify the crack profiles and derive the FPZ evolution formulations in different loading stages. According to the comparisons between the test results on the creep and control specimens, the effects of the preceding creep deformations during the creep tests on the crack initiation, the nonlinear fracture property, and the FPZ evolution of the rock-concrete interface were explored.

Section snippets

Materials and test specimens

The composite rock-concrete specimens were the beam specimens consisting of two identical rock and concrete blocks, and the pre-cracks were prepared at the bottom of the rock-concrete interface. Regarding the bilateral materials, the concrete with a strength grade of C40 and the granite produced in Liaoning Province of China were adopted in this study. The mix proportions of the concrete were 1: 0.60: 2.01: 3.74 (cement: water: sand: aggregate) by weight and the maximum aggregate size was

Determination of the crack profiles by using the DIC technique

By comparing the image at a loading point with the original one, the displacements at the pixel points in the loading image could be determined by the DIC analysis. In this study, no perfect descending branches of the P - CMODTPB curves were obtained in the TPB tests due to the significant brittleness of the rock-concrete interface. Seven loading stages, i.e. P1, P2, …, P7, were selected to conduct the DIC analysis, and the corresponding load values for individual loading stages are shown in

Crack initiation during the creep tests

At the beginning of the creep tests, the pre-crack did not propagate because the sustained load was smaller than the initial cracking load, i.e. Psus < Pini. With the increase of the loading time, the creep deformations developed gradually within the specimen, especially near the pre-crack tip due to the large local stress concentration [5]. The large stress concentration would lead to accumulations of micro-cracks near the pre-crack tip during the creep tests. However, it was uncertain whether

Conclusions

To investigate viscoelasticity-induced fracture behaviors of rock-concrete interface, creep tests were conducted on the composite rock-concrete specimens with two interface roughness degrees, i.e. 4 × 4 and 7 × 7 interfaces, under two sustained load levels of 50%Pmax and 75%Pmax. After the 90-day sustained loading, the creep specimens were unloaded and then reloaded to fail under quasi-static loading. During the reloading process, the DIC technique was employed to quantify the crack profiles

CRediT authorship contribution statement

Wenyan Yuan: Validation, Formal analysis, Investigation, Writing - original draft. Wei Dong: Conceptualization, Project administration, Supervision. Binsheng Zhang: Data curation, Writing - review & editing. Junzhou Huo: Writing - review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors gratefully acknowledge the financial support of the National Natural Science Foundation of China under the grants NSFC 51878117 and NSFC 52179123, and the Visiting Researcher Fund Program of State Key Laboratory of Water Resources and Hydropower Engineering Science under the grant 2020SGG02.

References (34)

Cited by (9)

View all citing articles on Scopus
View full text