This experimental study focuses on the links that exist between the mechanical pull-out behaviour of multi-filament yarns embedded in concrete and the impregnation of the yarn by the concrete. To this aim, 5 glass yarns have been embedded in concrete (AR and E glass yarns from assembled or direct roving). A pre-treatment was applied to the yarn before the casting, and was chosen in the following three: wetting, drying or pre-impregnation with a cement slurry. By this way, 15 yarn / pre-treatment combinations were obtained that generate 15 conditions of impregnation of the yarn by the cementitious matrix. In each case, the mechanical properties were determined from a classical pull-out test and the yarns impregnation properties were characterized by two dedicated tests: yarns porosity measurements and along yarn water flow measurements. By studying the relationship between the mechanical behaviour and the physical properties of the impregnated yarns, it is shown that the pre- and post-peak behaviour are mainly connected to the flow rate measured during the water flow measurements, which suggest a specific influence of the connected pores parallel to the filaments and of the penetration depth of the matrix into the yarn. The postpeak and the residual behaviour are moreover linked to the yarn pore volume associated to the disorder induced in the filaments assembly by the penetration of the matrix. The overall result of this work is a contribution to the understanding of the relationship between the impregnation of the yarns and the pull-out behaviour.

This experimental study focuses on the links that exist between the mechanical pull-out behaviour of multi-filament yarns embedded in concrete and the impregnation of the yarn by the concrete. To this aim, 5 glass yarns have been embedded in concrete (AR and E glass yarns from assembled or direct roving). A pre-treatment was applied to the yarn before the casting, and was chosen in the following three: wetting, drying or pre-impregnation with a cement slurry. By this way, 15 yarn / pre-treatment combinations were obtained that generate 15 conditions of impregnation of the yarn by the cementitious matrix. In each case, the mechanical properties were determined from a classical pull-out test and the yarns impregnation properties were characterized by two dedicated tests: yarns porosity measurements and along yarn water flow measurements. By studying the relationship between the mechanical behaviour and the physical properties of the impregnated yarns, it is shown that the pre- and post-peak behaviour are mainly connected to the flow rate measured during the water flow measurements, which suggest a specific influence of the connected pores parallel to the filaments and of the penetration depth of the matrix into the yarn. The postpeak and the residual behaviour are moreover linked to the yarn pore volume associated to the disorder induced in the filaments assembly by the penetration of the matrix. The overall result of this work is a contribution to the understanding of the relationship between the impregnation of the yarns and the pull-out behaviour.

Vortex breakdown is an important phenomenon observed in swirling flows involving the development of a stagnation point on the axis of the vortex followed by a region of recirculation when the swirl increases beyond a particular level. It has been studied extensively over past 50 years and various theories have been proposed to explain its various aspects. However, a single model explaining all the aspects together is yet to emerge. Numerical simulations of breakdown have been performed using a variety of grid-based as well as vortex methods. Vortex methods are a Lagrangian alternative to grid-based methods wherein the motion of the vorticity is determined by the local fluid velocity convection, with models for viscous effects when considered. The fluid velocity is obtained from the vorticity field. Only the rotational regions of the flow need to be considered leading to significant economy of computational effort for simulations of vorticity dominated flows, such as vortex breakdown. The inviscid vortex filament method has been used to simulate several aspects of the vortex breakdown phenomenon. The vortex filament method however, cannot easily simulate viscous effects. To simulate the viscous effects the viscous vortex particle method needs to be used. This work was intended to be a first step towards this end by initially evaluating the effectiveness of the inviscid version of the vortex particle method in simulating the breakdown phenomenon. The inviscid vortex particle method was found to satisfactorily simulate most qualitative aspects involved in the formation of vortex breakdown such as the retardation of axial velocity along centerline, radial swelling of the vortex core, formation of stagnation points, creation of azimuthal vorticity gradient from axial vorticity gradient and the turning of vortex lines along with the formation of a bubble-like structure with recirculating flow within. The effect of a wall placed adjacent to the vortex core was simulated by using image vortices. The wall was not found to influence the location of breakdown. However, the initiation of the spiral mode was found to occur earlier when a wall was present. For a quantitative assessment, a simulation of the experimental results of Faler and Leibovich (1978) was attempted. The simulation managed to predict the location of the breakdown and the extent of the bubble. The shape and height of the bubble obtained however were not in accord with the experimental observations. A single vortical cell was obtained in the interior of the bubble.

Zur Untersuchung der Schädigungs- und Versagensmechanismen eines in Beton eingebetteten Multifilamentgarns im Auszugversuch wurde die Schallemissionsanalyse zur Identifizierung und Lokalisierung von Filamentbrüchen eingesetzt. Im ersten Schritt wurden dazu die Schall emittierenden Ursachen (Filamentriss, Filamentablösung und Mikroriss im Beton) für eine Differenzierung charakterisiert. Es wurden Versuche zur Erzeugung von isolierten Signalen durchgeführt, welche mit Hilfe der Signal- und Frequenzanalyse untersucht wurden. Bei dem durchgeführten Garnauszugversuch konnte eine hohe Lokalisierungsgenauigkeit der Filamentbrüche erzielt werden. Der Schädigungsverlauf des Garns während des Auszugversuchs konnte detailliert untersucht werden.

Zur Charakterisierung der Mikrostruktur wurde die Geometrie des Verbundes aus REM-Aufnahmen ausgewertet, sowie die Verbundwirkung einzelner Filamente aus Filament-Pull-Out-Versuchen abgeleitet. Die gewonnenen Parameter bilden die Grundlage für die Modellierung von Garnauszugs- und Rissüberbrückungsmodellen. Unabhängig davon wurde ein Filament-Matrix-Verbund-Gesetz simuliert. Die Übereinstimmungen zwischen der Simulation und der experimentellen Untersuchung des Ausziehverhaltens werden in dem Artikel diskutiert. / The microstructure is characterized by analysing the bond geometrically and experimentally. The data to evaluate these two ways are from SEMimages and filament-pull-out-tests. Based on these data yarn-pull-out-models and crack-bridging-models are calculated. In addition a filament-matrix-bond-law is simulated independently. The correlation between the simulation and the experimental analysis of the pull-out behaviour is discussed in this paper.

Ces dernières années, les matériaux composites à matrices minérales ont suscité un grand intérêt dans le domaine du génie civil, en raison de leurs multiples avantages tels que la grande résistance en traction, le comportement pseudo-ductile, la facilité de mise en oeuvre, la nontoxicité et la compatibilité aux substrats. Toutefois, l’efficacité de ces composites est souvent limitée du fait de la faible adhérence du renfort avec la matrice. Cela est du à la faible pénétration des particules de la matrice entre les filaments internes du fil, ce qui entraine une rupture prématurée du matériau. Ce travail a pour but d’améliorer l’adhérence fil-matrice par pré-imprégnation du renfort, afin d’optimiser le comportement mécanique des composites. Deux méthodes de pré-imprégnation ont été utilisées : pré-imprégnation classique par voie humide et pré-imprégnation par voie sèche. Cette dernière se base sur l’application d’un champ électrostatique alternatif permettant la pénétration de la poudre au sein du renfort. La première méthode a déjà été utilisée dans la littérature et a permis d’améliorer les propriétés mécaniques du composite. Cependant, son utilisation reste limitée aux éléments préfabriqués simples contrairement à la deuxième méthode qui peut être appliquée pour réaliser des éléments de construction de grande taille, ou bien pour le renforcement et la réparation des structures in situ. La présente étude s’appuie sur une analyse multi-échelles permettant d’aborder le problème d’adhérence à l’échelle d’interface (fil-matrice) et à l’échelle du composite (fil-matrice et textile-matrice). Tout d’abord, des essais d’arrachement ont été conduits sur des fils de verre enrobés dans une matrice ettringitique. Huit configurations ont été étudiées afin d’évaluer l’effet du processus de pré-imprégnation et le type des matériaux de pré-imprégnation à sec. Il en sort que le comportement d’arrachement est amélioré par pré-imprégnation du fil par voie humide et sèche mais, précisons que par voie sèche, l’amélioration est significative seulement dans le cas d’utilisation de poudre de particules très fines. Une modélisation numérique de l’essai d’arrachement a été également réalisée afin d’identifier les propriétés d’interface fil-matrice. En se basant sur ces résultats, une étude expérimentale et numérique du comportement en traction des composites fil-matrice a été ensuite réalisée. Cinq configurations ont été retenues, après avoir écarté les prétraitements de fil ayant eu un effet néfaste sur l’adhérence fil-matrice. Il s’est avéré que les deux procédés contribuent à une augmentation significative des propriétés mécaniques du composite. De plus, un modèle numérique 3D a été développé pour simuler l’essai de traction d’un composite sec et pré-imprégné. Il a permis d’une part de prédire le comportement global du matériau et d’autre part d’apporter une meilleure compréhension de ce comportement grâce à l’analyse des états de déformation et de contrainte au niveau du fil, de la matrice et de l’interface. Enfin, une comparaison de l’effet des deux procédés de pré-imprégnation a été établie à l’échelle du composite textile-matrice au moyen d’essais de traction et d’analyse de corrélation d’image. Cinq différents types de tissu de verre unidirectionnel ont été utilisés. La pré-imprégnation par voie sèche a démontré la plus forte amélioration des performances du composite quel que soit le grammage du tissu, contrairement à la pré-imprégnation par voie humide qui trouve sa limite quand le textile devient dense / In recent years, mineral matrix composites have gained great interest in the field of civil engineering, thanks to their multiple advantages such as high tensile strength, pseudo-ductile behavior, ease of implementation, non-toxicity and substrate compatibility. However, the efficiency of these composites is often limited because of the low adhesion of the reinforcement to the matrix. This is due to the low penetrability of matrix particles into the space between the inner filaments of the yarn, which leads to premature failure of the material. The purpose of this work is to improve the yarn-matrix adhesion by pre-impregnating the reinforcement in order to optimize the mechanical behavior of composites. Two pre-impregnation methods were used : conventional pre-impregnation in a wet manner and pre-impregnation in a dry manner. This last is based on an alternating electrostatic field used to impregnate yarns with powder. The first method has already been used in the literature and has improved the mechanical properties of the composite. However, it is suited only for the manufacturing of simple prefabricated elements, unlike the second method, which can be applied for the construction of large elements, or for strengthening or repairing structures in site. This study is based on a multi-scale analysis to investigate the bond problem at the interface scale (yarn-matrix) and at the composite scale (yarn-matrix and textile-matrix). First, pull-out tests were conducted on glass yarns embedded in an ettringitic matrix. Eight configurations were studied to evaluate the effect of the pre-impregnation process and the type of dry pre-impregnation powders. It results that the pull-out behavior is improved by pre-impregnating the yarn in wet and dry manner but, it should be noted that in dry process, the improvement is significant only when using very fine particle powders. A numerical modeling of the pull-out test was also carried out in order to identify the yarn-matrix interface properties. Based on these results, an experimental and numerical study of the tensile behavior of yarn-matrix composites was then carried out. Five configurations were selected, after excluding the yarn pre-treatments that had a negative effect on the yarn-matrix bond. It was found that both processes contribute to a significant increase in the mechanical properties of the composite. In addition, a 3D numerical model has been developed to simulate the tensile test of a dry and pre-impregnated composite. It allowed on the one hand to predict the global behavior of the material and on the other hand to provide a better understanding of this behavior by analyzing the deformation and stress states at the level of the yarn, the matrix and the interface. Finally, a comparison of the effect of the two pre-impregnation processes was established at the textile-matrix composite scale by means of tensile tests and image correlation analysis. Five different types of unidirectional glass fabric were used. The dry pre-impregnation has shown the best improvement in composite performance regardless of fabric density, unlike the wet pre-impregnation, which has its limits when the textile becomes dense

Zur Charakterisierung der Mikrostruktur wurde die Geometrie des Verbundes aus REM-Aufnahmen ausgewertet, sowie die Verbundwirkung einzelner Filamente aus Filament-Pull-Out-Versuchen abgeleitet. Die gewonnenen Parameter bilden die Grundlage für die Modellierung von Garnauszugs- und Rissüberbrückungsmodellen. Unabhängig davon wurde ein Filament-Matrix-Verbund-Gesetz simuliert. Die Übereinstimmungen zwischen der Simulation und der experimentellen Untersuchung des Ausziehverhaltens werden in dem Artikel diskutiert. / The microstructure is characterized by analysing the bond geometrically and experimentally. The data to evaluate these two ways are from SEMimages and filament-pull-out-tests. Based on these data yarn-pull-out-models and crack-bridging-models are calculated. In addition a filament-matrix-bond-law is simulated independently. The correlation between the simulation and the experimental analysis of the pull-out behaviour is discussed in this paper.

Zur Untersuchung der Schädigungs- und Versagensmechanismen eines in Beton eingebetteten Multifilamentgarns im Auszugversuch wurde die Schallemissionsanalyse zur Identifizierung und Lokalisierung von Filamentbrüchen eingesetzt. Im ersten Schritt wurden dazu die Schall emittierenden Ursachen (Filamentriss, Filamentablösung und Mikroriss im Beton) für eine Differenzierung charakterisiert. Es wurden Versuche zur Erzeugung von isolierten Signalen durchgeführt, welche mit Hilfe der Signal- und Frequenzanalyse untersucht wurden. Bei dem durchgeführten Garnauszugversuch konnte eine hohe Lokalisierungsgenauigkeit der Filamentbrüche erzielt werden. Der Schädigungsverlauf des Garns während des Auszugversuchs konnte detailliert untersucht werden.

Cette recherche porte sur les fils multifilamentaires de verre utilisés pour renforcer les matériaux à base de ciment. Elle est focalisée sur les interactions mécaniques de ce type de fils, constitués d'un assemblage de milliers de filaments micrométriques, avec un micro-béton et sur le rôle spécifique de l'imprégnation du fil par cette matrice cimentaire. Trois pré-conditionnements des fils ont été employés lors de la fabrication des éprouvettes afin de moduler les conditions d'imprégnation par la matrice cimentaire. L'imprégnation de 5 fils multi-filamentaires par la matrice cimentaire a été caractérisée et les paramètres d'imprégnation ont été définis en s'appuyant sur des observations MEB, ainsi que des essais de porosimétrie au mercure et des essais spécifiques d'écoulement le long du fil enrobé. Des essais classiques d'arrachement de type pull-out ont été utilisés pour la caractérisation mécanique. L'analyse des liens entre les propriétés mécaniques et les paramètres d'imprégnation ont permis de mieux comprendre les micro-mécanismes d'interaction filaments / matrice cimentaire et d'expliquer le comportement macroscopique à l'arrachement. / This research deals with multi-filaments glass yarns used as reinforcement of cement based materials. It focuses on the mechanical interactions of these yarns, made of thousands of micrometric filaments, with a micro-concrete and on the specific part of the impregnation of the yarn by the cementitious matrix. Modulated impregnation conditions of the yarns were obtained by using three different manufacturing processes for the samples preparation. The impregnation of 5 multi-filament yarns by the cementitious matrix has been characterized and physical parameters of impregnation were determined using SEM investigations, mercury intrusion porosity measurements and specific tests of flow all along the embedded yarn. Classical pull-out tests have been used for the mechanical characterisation. The study of the links between the mechanical properties and the physical parameters of impregnation allowed accessing a better understanding of the filaments / cementitious matrix interaction micro-mechanisms, and explaining the macroscopic pull-out behaviour.

The paper presents a meso-scopic modeling framework for the simulation of three-phase composite consisting of a brittle cementitious matrix and reinforcing AR-glass yarns impregnated with epoxy resin. The construction of the model is closely related to the experimental program covering both the meso-scale test (yarn tensile test and double sided pull-out test) and the macro-scale test in the form of tensile test on the textile reinforced concrete specimen. The predictions obtained using the model are validated using a-posteriori performed experiments.