Academic literature on the topic 'Quasi-brittle geomaterials'

Nonlocal criteria are used for prediction materials and rock mass failure near stress concentrations (pores, faults, openings, excavations). A common property of nonlocal fracture criteria is the introduction of the intrinsic material length characterizing its microstructure, which allows one to describe the size effect in conditions of stress concentration. At the same time the scope of their application is limited to cases of brittle or quasi-brittle fracture with a small fracture process zone. To expand the scope of the criteria for cases of fracture with a developed fracture process zone, it is proposed to abandon the hypothesis of the size of this zone as a material constant, associated only with the material structure. New fracture criteria are proposed, which are the development of the average stress criterion, and point stress criterion, and which contain a complex parameter that characterizes the size of the fracture process zone and accounts not only for the material structure, but also plastic properties of the material, geometry of the sample, and its loading conditions. Expressions are obtained for the critical pressure in the problem of the formation of tensile cracks under compression in the samples of geomaterials with a circular hole. The calculation results are in good agreement with the experimental data on the fracture of drilled gypsum plates.

The study is aimed at the development of the new failure criteria for quasi-brittle materials in conditions of stress concentration. The possibility of using non-local failure criteria for description of the brittle, quasi-brittle and ductile fracture of the materials with notches is analyzed. The general feature of these criteria consists in the introduction of the internal dimension characterizing the structure of the material, which provides the possibility of describing a large-scale effect in conditions of the stress concentration and thereby expand the area of their application compared to traditional criteria though it is limited to the cases of brittle or quasi-brittle fracture with a small pre-ffacture zone. To broaden the scope of their application to quasi-brittle fracture with a developed pre-fracture zone we propose to abandon the hypothesis about the size of the pre-fracture zone as a constant related only to the structure of the material. A number of the new nonlocal criteria, which are the development of the criteria of the mean stress and fictitious crack, are developed, substantiated from the physical standpoint, and proved experimentally. These criteria contain a complex parameter characterizing the size of the pre-fracture zone and taking into account not only the structure, but also the ductile properties of the material, specimen geometry and loading conditions. The expressions for the critical pressure in the problem of tensile crack formation upon compression of the samples of geomaterials with a circular hole are derived. The results of calculations match rather well the experimental data on the destruction of drilled gypsum slabs.

Fracture under compression is one of the most commonly studied properties of geomaterials like concrete and rock, in particular since these materials reach their best performance in compression. The fracture process is however rather complex due to the heterogeneous structures of said materials. Over the years fundemental studies of fracture under compression have led to a much improved insight in the details of the fracture process depending on the actual composition of the material. Fracture can be described by means of a 4-stage fracture model, which included as most important aspects pre-peak cracking, which is stable and can be arrested by stiffer and stronger elements in the material structure, and post-peak cracking [1]. The latter macroscopic cracks are basically un-stable and can only be arrested by measures at a structural scale, such as applying confining stress or by using positive geometries. The material structure cannot assist in the arrest of the large energetic cracks other than locally affecting the crack path. In the paper an overview is given of the fracture process in compression. Recently we embarked on studying compressive fracture using a simpler material structure, namely foamed hardened cement paste [2]. Stiff aggregates that are normally included in normal concrete have been left-out; instead a larger than usual quantity of large pores is brought into the material, even up to 80%. Studying fracture processes in this simpler material system ultimately allows for a better understanding of the details of the pre-peak cracking process, which is considered more important than the post-peak process since it defines strength.

Propagation des fissures microscopiques, est représentée par des variables d’endommagement. L’évolution de la variable d’endommagement est généralement formulée sur la base d’observations expérimentales. De nombreux modèles phénoménologiques d’endommagement ont été proposés dans la littérature. L’objet de cette thèse est de développer une nouvelle procédure pour obtenir des lois d’évolution macroscopique d’endommagement,dans lesquelles l’évolution de l’endommagement est entièrement déduite de l’analyse de la microstructure. Nous utilisons une homogénéisation basée sur des développements asymptotiques pour décrire le comportement global à partir de la description explicite d’un volume élémentaire microfissuré.Nous considérons d’une part un critère quasi-fragile (indépendant du temps) puis un critère sous-critique(dépendant du temps) pour décrire la propagation des microfissures. De plus, le frottement entre les lèvres des microfissures est pris en compte. Une analyse énergétique est proposée, conduisant à une loi d’évolution d’endommagement qui intègre une dégradation de la rigidité, un adoucissement du comportement du matériau, des effets de taille et d’unilatéralité, mettant en avant un comportement différent à la rupture en contact avec et sans frottement. L’information sur les micro-fissures est contenue dans les coefficients homogénéisés et dans la loi d’évolution de l’endommagement. Les coefficients homogénéisés décrivent la réponse globale en présence de micro-fissures (éventuellement statiques), tels qu’ils sont calculées avec la(quasi-) solution microscopique statique. La loi d’endommagement contient l’information sur l’évolution des micro-fissures, résultant de l’équilibre énergétique dans le temps pendant la propagation microscopique.La loi homogénéisée est formulée en incrément de contrainte. Les coefficients homogénéisés sont calculées numériquement pour des longueurs de fissures et des orientations différentes. Cela permet la construction complète des lois macroscopiques. Une première analyse concerne le comportement local macroscopique, pour des trajets de chargement complexes, afin de comprendre le comportement prédit par le modèle à deux échelles et l’influence des paramètres micro structuraux, comme par exemple le coefficient de frottement. Ensuite, la mise en œuvre en éléments finis des équations macroscopiques est effectuée et des simulations pour différents essais de compression sont réalisées. Les résultats des simulations numériques sont comparés avec les résultats expérimentaux obtenus en utilisant un nouvel appareil triaxial récemment mis au point au Laboratoire 3SR à Grenoble (France)
In continuum damage models, the degradation of the elastic moduli, as the results of microscopic crackgrowth, is represented through damage variables. The evolution of damage variable is generally postulatedbased on the results of the experimental observations. Many such phenomenological damage modelshave been proposed in the literature. The purpose of this contribution is to develop a new procedurein order to obtain macroscopic damage evolution laws, in which the damage evolution is completelydeduced from micro-structural analysis. We use homogenization based on two-scale asymptotic developmentsto describe the overall behaviour starting from explicit description of elementary volumes withmicro-cracks. We consider quasi-brittle (time independent) and sub-critical (time dependent) criteria formicro-cracks propagation. Additionally, frictional contact is assumed on the crack faces. An appropriatemicro-mechanical energy analysis is proposed, leading to a damage evolution law that incorporates stiffnessdegradation, material softening, size effect, and unilaterality, different fracture behaviour in contactwithout and with friction. The information about micro-cracks is contained in the homogenized coefficientsand in the damage evolution law. The homogenized coefficients describe the overall response inthe presence of (possibly static) micro-cracks, as they are computed with the (quasi-) static microscopicsolution. The damage law contains the information about the evolution of micro-cracks, as a result ofthe energy balance in time during the microscopic propagation. The homogenized law is obtained in therate form. Effective coefficients are numerically computed for different crack lengths and orientations.This allows for the complete construction of the macroscopic laws. A first analysis concerns the localmacroscopic behaviour, for complex loading paths, in order to understand the behaviour predicted bythe two-scale model and the influence of micro structural parameters, like for example friction coefficient.Next, the FEM implementation of the macroscopic equations is performed and simulations for variouscompression tests are conducted. The results of the numerical simulations are compared with the experimentalresults obtained using a new true-triaxial apparatus recently developed at the Laboratory 3SRin Grenoble (France)

The formation and development of localisation bands and/or cracks have been experimentally identified as the key failure mechanism governing responses of quasi-brittle geomaterials like concrete, sandstones and soft rocks. In fact, key features of the material behaviour, including Lode-angle dependence, size effect and brittle-ductile transition can be considered as consequences stemming from the localised failure mechanism in several loading conditions. For geomaterials with fibre reinforcement (i.e., fibre reinforced concrete), even though adding short fibres into the material significantly changes its mechanical characteristics and performances, the development of cracks and localisation bands remains the central mechanism driving the material responses. In this case, the bridging effect caused by fibres across cracks, together with the material cohesive-frictional resistance, refrains cracks from further developing and forces the material to form more cracks throughout the structure to dissipate the given energy. This prolongs the coalescence of diffuse micro/meso-cracks to form a macro-crack and considerably improve the material strength and ductility. Classical continuum models, in principle, can capture the overall responses of the material with stress-strain relationships formulated from experimental observations at the macroscopic level. However, the material behaviour in these models is homogenous throughout the whole element domain, leading to an incorrect dependence of dissipated energy and specimen responses on the discretisation resolution. This is because they fail to capture the difference of deformation and behaviour between the localisation zone and its surrounding bulk material. As a result, with the presence of crack/localisation band, the definition of averaged quantities such as overall stress and strain is not adequate to describe the volume element and using them for analysing post-localisation behaviour of quasi-brittle geomaterials is inappropriate, if not totally incorrect. Consequently, classical approaches that ignore the strong heterogeneity induced by the localisation of deformation suffer from mesh convergence issues in Boundary Value Problems (BVPs) analysis. In this research, the localised failure mechanism is employed as the basis to develop a continuum-based constitutive model for quasi-brittle geomaterials with and without fibre reinforcement. The cracks/localisation bands are explicitly included as an intrinsic part of the model with its own behaviour in conjunction with the responses of the surrounding bulk material. This allows an additional constitutive relationship, together with its internal variables, to be defined inside the localisation band to describe its microstructural changes under the course of loading. The material inelastic behaviour and important features such as brittle-ductile transition, Lode-angle dependence, size effect and hydrostatic pressure dependence can thus be correctly captured in association with observable failure patterns at constitutive level. The model, constructed in this manner, is also capable of featuring multiple localisation bands/cracks inside the constitutive equations to account for any change of loading path and avoid unphysical stress-locking naturally. In addition, for modelling quasi-brittle geomaterials with fibre reinforcement, the incorporation of cracks within the constitutive model enables the inclusion of separate models describing the fibre bridging effect and material cohesive-frictional resistance. As a result, the interactions between these two components and their contributions to the stress transfer across a crack are naturally captured for different types of fibres and their volume contents. The transition from hardening to softening, corresponding to the change from diffuse cracking phase to localised failure can also be reflected as a consequence. Furthermore, owing to the explicit inclusion of cracks in the model, the resulting constitutive behaviour automatically scales with the discretisation resolution and the results are thus mesh-independent when solving BVPs, without requiring any additional regularisation. Model validations against experimental data show that the proposed model is simple yet effective in capturing the localised failure of quasi-brittle geomaterials with and without fibre reinforcement at both constitutive and structural levels. The model is proven to be reliable and computationally inexpensive, with a few model parameters which can be identified and calibrated in a consistent and physically meaningful manner. The proposed model can thus be applied straightforwardly for the analysis and design of solids/structures made of rocks or concrete, with or without fibre reinforcements.
Thesis (Ph.D.) -- University of Adelaide, School of Civil, Environmental and Mining Engineering, 2019