Disc brake calipers are subjected to complex mechanical loading and interaction of individual components in a typical brake assembly makes design improvement very challenging. To analyse caliper behaviour, complex Finite Element models were created and successfully validated using a variety of experimental techniques, including exceptionally suitable Digital Image Correlation. A novel methodology to optimise caliper design was developed, using non-linear contact Finite Element Analysis and topology optimisation, to generate lightweight, high performance brake calipers. The method was used on a Formula 1 brake assembly and significant improvement in structural design was achieved, with the new caliper being lighter and stiffer than the original. The same approach was used on more conventional 4 pistons calipers using various boundary conditions with particular focus on mass reduction and considerably lighter designs were achieved. The influence of specific features of the optimised calipers on the structural performance was also successfully investigated.

Titanium is a widely used metal in industrial and commercial applications. It retains anisotropic mechanical properties at room temperature due to its HCP crystal structure. The effects of crystal orientation have been studied theoretically and through modeling though there is a lack of empirical data available on the topic. The work presented here uses laser-machined voids along with EBSD analysis to study the ductility of grains in different orientations to better understand the microscale fracture process in α-titanium. Experimental results show that hard grains with their c-axis parallel to the tensile direction behave in a less ductile manner than grains with their c-axis oriented away from the tensile direction. This is due to the basal slip systems activating in the former case and prismatic slip systems in the latter. Models utilized include the McClintock model for void growth, Brown-Embury model for void coalescence and FEM crystal plasticity simulations

A pipe consisting of Inconel 600 welded to grade 106-B Carbon-Steel using Inconel 182 weld filler is used to transport heavy water in nuclear reactors. A confidential report concluded that cracking is one of the problems these pipes are currently facing. Before cracking can be fully understood the mechanical properties of the weld must be determined. This thesis analyzed the pipe at various length-scales using optical microscopy, micro-hardness testing, small and large scale tensile testing and digital image correlation (DIC). This thesis successfully achieved it goals of determining the mechanical properties and creating a model of the Inconel dissimilar metal weld. It partially met the goal of observing fracture mechanisms as it was able to observe fracture in tensile samples but was not able to successfully track crack growth.

Une méthode de corrélation d'images numériques adaptée aux mesures cinématiques dans les polycristaux a été développée dans ce travail, dans le but d'identifier les paramètres de lois de plasticité cristalline. Des mesures de champs 2D sont réalisées à la surface d'un polycristal d'acier austénitique 316LN, à partir d'une série d'images acquises au Microscope Electronique à Balayage (MEB) au cours d'un essai de traction monotone in-situ, et pour différentes tailles moyennes de grains. Pour permettre la corrélation d'images, un mouchetis adapté à l'échelle microscopique est au préalable déposé à la surface de l'éprouvette par une technique de microlithographie. Les distorsions spatiales liées à l'imagerie MEB et à la technique de marquage sont quantifiées. La connaissance de la microstructure en surface par diffraction des électrons rétrodiffusés permet de réaliser les mesures cinématiques en utilisant un maillage éléments finis non-structurés s'appuyant sur les joints de grains ou de macles. Ce même maillage est ensuite utilisé pour la simulation de chaque essai de traction sur la microstructure expérimentale, avec comme conditions aux limites les déplacements nodaux mesurés au cours du temps sur les bords du domaine. Deux lois de plasticité cristalline d'approche locale sont alors considérées pour simuler les hétérogénéités des déformations observées : la loi de Méric-Cailletaud et la loi DD-CFC développée à EDF R&D et davantage basée sur la physique. Des comparaisons entre la mesure et la simulation sont menées, en termes de déplacements, de déformations, mais aussi de systèmes de glissement activés. Finalement, une méthode inverse d'identification des paramètres de loi est proposée, se basant sur le recalage à la fois des champs de déplacements locaux et du comportement homogène du matériau. Les paramètres de l'écrouissage isotrope de la loi de Méric-Cailletaud sont ainsi identifiés pour différentes tailles moyennes de grains. Il est également montré que certain coefficients de la matrice d'interaction entre systèmes de glissement peuvent être estimés. / A digital image correlation procedure adapted to kinematic measurements in polycrystals has been developed in this work to identify parameters of crystal plasticity laws. 2D kinematic measurements are performed on the surface of 316LN austenitic steel polycrystals from a sequence of images acquired using a Scanning Electron Microscope (SEM) during in-situ tensile tests for various mean grain sizes. To enable digital image correlation, a speckle adapted to the microscopic scale is deposited onto the specimen surface by a microlithography process. Spatial distortions resulting from both patterning and SEM imaging techniques are quantified. The knowledge of the microstructure at the surface by electron backscattered diffraction allows for kinematic measurements to be performed using an unstructured finite element mesh taking as support the grain or twin boundaries. This same mesh is then used for the simulation of each tensile test on the experimental microstructure with the measured nodal displacements prescribed as boundary conditions with their time evolution. Two local crystal plasticity laws are considered to simulate the observed strain heterogeneities, namely, the Méric-Cailletaud model and the DD-CFC law developed at EDF R&D. Comparisons between measurements and simulations are performed in terms of displacements, strains but also activated slip systems. Last, an inverse identification method is proposed for the identification of the sought constitutive parameters based on both the local displacement fields and the material homogenized behavior. The parameters associated with isotropic hardening of Méric-Cailletaud law are thus identified for various mean grain sizes. It is also shown that some of the interaction parameters of slip systems can be estimated.

L'objectif de cette recherche est d’analyser l’apparition et la propagation des fissures de dessiccation et de mieux comprendre la relation entre les comportements macroscopique et microscopique de cinq argiles, une kaolinite, une montmorillonite et trois mélanges de kaolinite et montmorillonite. A l'échelle macroscopique, la méthode est basée sur (1) la mesure de la teneur en eau, de l’indice des vides et du degré de saturation en fonction de la succion au cours du séchage, qui permet de préciser la relation entre le retrait et la désaturation et met en évidence les phases caractéristiques du comportement; (2) des mesures de teneur en eau et des déformations globales dans les tests de dessiccation libre afin d'étudier leur homogénéité ; (3) la détermination des déformations et des déplacements locaux pendant le séchage avec les logiciels VIC-2D et VIC-3D ; (4) une étude classique des paramètres des fissures ; (5) des essais de traction pour déterminer les propriétés des argiles impliquées dans la formation de fissures . A l'échelle microscopique, l'étude est basée sur une analyse approfondie de la microstructure en utilisant en particulier le microscope électronique à balayage, couplé à une méthode adaptée pour l'identification des orientations préférentielles des particules. Cette étude microscopique est complétée par l'analyse de porosimétrie au mercure, une méthode qui permet de quantifier l'espace poral et de caractériser l’indice des vides local. Outre les recherches sur les sols argileux, les effets de la décompression et de la succion sur la formation des fissures dans une roche argileuse ont été analysés également. La relation entre les changements macroscopiques, les changements dans la microstructure et de la porosité a été étudiées. Lors des essais de dessiccation libre, les déformations et les déplacements sont obtenus avec Vic-2D. Les zones de l’échantillon où les fissures apparaissent sont identifiées ainsi que l’évolution des déformations et des déplacements avant l’apparition des fissures. A la fin du séchage, les fissures forment une sorte de réseau, où des bifurcation peuvent être observées dans certains cas. Deux modes de fissuration sont détectés pendant les essais : le mode de traction et le mode de déchirement. Lorsqu’une fissure est causée par des tractions, la direction de propagation suit la direction perpendiculaire à l’extension maximale. S’il y a des distorsions à proximité de la fissure, alors sa direction change. Au début du séchage, les déplacements et les déformations sur le bord sont plus importants que dans les autres parties du modèle. Au voisinage des fissures, les déplacements et les déformations deviennent plus grands que dans les autres parties. Dans la plus grande partie du modèle, les déformations principales sont essentiellement longitudinales et transversales. Pour un matériau donné, la résistance à la traction augmente lorsque la teneur en eau diminue. Le tracé des résultats en fonction de l’indice de liquidité permet de mettre en évidence l’effet de la minéralogie sur la résistance à la traction. Au même indice de liquidité, la contrainte maximale de traction diminue lorsque le pourcentage de monmorillonite augmente. L’effet de la succion sur la résistance à la traction semble qualitativement similaire à ce qui est observé dans le cas de la résistance à la compression simple : on observe que la résistance à la traction est une fonction linéaire du logarithme de la succion. L’analyse au MEB de quatre suspensions soumises à plusieurs succions met en évidence l’isotropie globale de la microstructure, avec une orientation aléatoire des particules, tandis qu’une analyse plus fine révèle que la microstructure peut présenter localement une certaine anisotropie. / The objective of this research is to analyze the appearance and propagation of cracks related to desiccation and to provide a better understanding of the relation between the macroscopic and microscopic behavior of five clays, a kaolinite, a montmorillonite and three mixtures of kaolinite and montmorillonite. At the macroscopic scale, the method is based on (1) measurements of water content, void ratio and degree of saturation versus suction during drying, which allows to specify the relationship between shrinkage and desaturation and highlights the characteristic phases of behavior; (2) measurements of water contents and global deformations in free desiccation tests in order to study their homogeneity; (3) the determination of the local deformations and displacements during drying using the softwares VIC-2D and VIC-3D; (4) a classical study of the parameters of cracks; (5) traction tests in order to identify the tensile properties of the clays involved in the formation of cracks. At the microscopic scale, the study is based on a thorough microstructure analysis using in particular scanning electron microscope, coupled to an adapted method for the identification of preferential orientations of particles. This microscopic study is complemented by mercury intrusion porosimetry analysis, a method which allows to quantify the poral space and to characterize the local void ratio. In addition to the research on clay soils, the effects of decompression and suction on the formation of cracks in a clay rock were analyzed too. The relationship between macroscopic changes and the changes in the microstructure and porosity was investigated. During free desiccation tests, two-dimensional strains and displacements maps are obtained with Vic-2D. The zones of the sample where cracks appear are identified as well as the evolution of strains and displacements before the appearance of cracks. At the end of desiccation, the cracks form a kind of network. Bifurcation of cracks can be observed in some cases. Two modes of cracks are detected during the tests: traction mode and tearing mode. When a crack is caused by traction, the propagation direction follows the direction perpendicular to maximum extension. If there are shear strains in the vicinity of the crack, then its direction changes. In the early time of desiccation, the displacements and strains on the boundaries are larger than those in the other parts of the model. In the vicinity of cracks, displacements and strains are relatively larger than those in the other parts. In most parts of the model, the principal strains are mainly longitudinal and transversal. For a given material, tensile strength increases when water content decreases. Plotting the results versus the liquidity index allows highlighting the effect of mineralogy on tensile strength. At the same liquidity index, the maximum tensile stress decreases when the montmorillonite content increases. The effect of suction on tensile strength seems qualitatively similar to what is observed in the case of unconfined compression strength with tensile strength being a linear function of the logarithm of suction. Analysis with SEM of four slurries submitted to several suctions highlights the global isotropy of the microfabric, with a random orientation of the particles, while a finer analysis reveals that the fabric may present locally some anisotropy.

Being able to predict the evolution of plastic deformation at the microstructural scale is of paramount importance in the engineering of materials for advanced applications. However, this is not straightforward because of the multiscale nature of deformation heterogeneity, both in space and time . The present thesis combines four related studies in a coherent work, which is aimed to develop experimental methods for studying crystal plasticity at the micro and mesoscale. A novel methodology for gold remodelling is initially proposed and used to apply high-density speckle patterns on the surface of stainless steel specimens. The unique proprieties of the speckle pattern enabled plastic deformation mapping with submicron resolution using digital image correlation (HDIC). It was therefore possible to study the concomitant evolution of microbands and transgranular deformation bands in such alloy. High-resolution deformation mapping also enabled comparison with high-resolution electron backscatter diffraction (EBSD) observations. The only partial correspondence of results proved the limits of EBSD in characterizing plastic deformation. The cause of such limitation is later identified in the reduced sensitivity to lattice slip of the EBSD technique. Hence, a novel method of HDIC data analysis is proposed to separate the contributions of lattice slip and lattice rotation from the deformation mapping. The method is adopted to characterize plasticity in austenitic stainless steel and at the plastic deformation zone (PDZ) around a silicon particle embedded in a softer aluminum matrix. Results show that the proposed experimental methodology has the unique capability of providing a complete description of the micro and mesoscale mechanics of crystal plasticity. HDIC therefore emerges as a key technique in the development of accurate physical-based multiscale crystal plasticity models.

It has been found that commercially pure (CP) Titanium (Ti) undergoes substantial amount of deformation twinning during plastic forming in a wide range of temperatures and strain rates giving CP-Ti good ductility and allowing up to 90% thickness reduction by cold rolling. Aluminium (Al) rich ! Ti-alloys lack this superior ductility but exhibit therefore up to five times higher yield strength, which was connected experimentally to reduced activity of deformation twinning with addition of Al to Ti. Ultimately this is also valid in the ! phase of two-phase alloys such as Ti6Al4V and thought to be key to the reduced ductility in Al rich alloys. It is to date unclear if ordering of Al in the Ti matrix, a change in the stacking fault energy (SFE) with alloying or a transition of the cellular dislocation structures in CP-Ti to planar slip patterns at high Al contents reduces twin activity. The focus of this dissertation project is therefore the transition of microstructural details and the deformation structures in the ! phase with increasing Al concentration. For simplified investigations binary Ti-Al alloys containing 3.5, 7, 10 and 13at.% Al have been created with comparable grain morphology and texture within this study. For a better understanding of the role of Al also binary Ti-Sn (Tin) alloys (1 & 3.4at.% Sn) and Ti-Zr (Zirconium) alloys (3.6 & 10at.%) as well as an Oxygen (O) rich Ti-10at.%Al and the industrial compositions of Ti6Al4V were produced on the same route and investigated by the same methods. This alloy range allows evaluation of the effects of the c/a ratio, ordering phenomena and the SFE on the twin activity. The knowledge was finally transferred to industrially forged CP-Ti and Ti5Al2.5Sn. TEM and neutron diffraction confirmed the onset of Ti3Al formation from Al concentrations above 7at.% (4wt%), but no ordering of Zr or Sn atoms was found after solution treatments. The evolution of lattice strain and lattice reorientation due to twinning with increasing compressive strain was captured by in-situ experiments under neutron diffraction at Engin-X, ISIS. Post-mortem EBSD micro and macro texture mappings revealed that the twin fraction in Al reduces above a critical concentration of 7at.% (4wt%), which was enhanced with increasing ordering towards Ti3Al. Sn and Zr addition showed no significant effect on the overall twin fraction, but increased twin numbers with facilitated nucleation and impeded twin growth, which may be related to the SFE. Increasing slip planarity and a transition from prismatic slip towards basal slip with addition of Al was found with means of Digital image correlation (DIC). DIC also revealed intense prismatic slip in grains undergoing !"!! tension twinning and virtually barely any strain accumulation within a twin below 9% plastic strain, rationalised by much increased nanohardness in the twin in comparison to the parent. Nanoindentation also revealed that alloying with Al reduces the crystal anisotropy. Finally it is believed that ordering and the closely related transition of slip patterns lead to the reduction in twin activity, while c/a ratio, crystal anisotropy and SFE seem less important.

Except for aerodynamics forces, all vehicle excitation forces are generated at the tyre-road interface. Considering low speed applications, such as terramechanics where the aerodynamics force are negligible, the road-tyre interaction is of extreme importance. Crucial variables which govern the forces generated at the tyre-road interface are tyre side-slip angle, tyre longitudinal slip ratio and terrain pro le. Solutions to measure these variables exist for smooth hard roads but the solutions experience challenges on rough and o -road terrain. Digital Image Correlation is concerned with tracking the changes of a scene in a sequence of images or in images obtained from multiple viewpoints. These methods are frequently used in micro and nano-scale mechanical testing due to its ease of implementation and use as well as its non-contact approach. As a result these techniques are being implemented in many elds from material testing, physics, lm animations and engineering. The aim of this thesis is to investigate the applications of Digital Image Correlation in vehicle dynamics using cost e cient o -the-shelf digital cameras and lenses. The following three vehicle dynamic problems are investigated: The measurement of the vehicle side-slip angle, longitudinal slip-ratio of a pneumatic tyre and high delity terrain pro ling. The vehicle side-slip angle can be used as a measure of the vehicle stability and therefore be used to improve the e ectiveness of vehicle stability controllers. The tyre side-slip angle is also a vital measurement in characterising the lateral force characteristics of pneumatic tyres. A planar measuring method using Digital Image Correlation is shown to accurately measure the side-slip angle. The method is expanded by developing two additional algorithms which can measure all translational and rotational velocities. These methods are validated on both smooth surfaces and rough o -road terrain. A method is also implemented whereby the longitudinal slip ratio of a tyre can be measured using a single camera. Therefore, doing away with the conventional method of using three independent measurement systems. Features in the contact patch, encompassing both the tyre and the road, are tracked in a sequence of images. The features are classi ed into features lying on the tyre, road and outliers using a clustering algorithm. This enables the system to determine the tyre and road velocities from which the slip ratio is determined. High delity terrain pro ling is performed using a calibrated stereographic rig to obtain a three dimensional point cloud of the scene which is being viewed. The point cloud generated at one sample contains a grid of points encompassing a large area with points spaced both laterally and longitudinally. Overlapping point clouds are generated and joined using various registration techniques. The joined point clouds are sub-sampled to obtain a regularised grid of point containing a single point cloud of non overlapping points. The proposed techniques create new possibilities in the eld of vehicle dynamics. Enabling the side-slip angle to be measured in rough of road conditions while providing additional measurements. The longitudinal slip ratio which is measured directly at the contact patch could pave the way for better understanding the mechanism of the longitudinal tyre force generation. The inexpensive road pro ling systems enables multiple sensors to be used in terramechanics tests to determine the impact of a vehicle on the environment. The thesis presents the mere tip of the ice berg concerning digital image correlation used in vehicle dynamics with many more possibilities waiting to be discovered. / Thesis (PhD)--University of Pretoria, 2015. / tm2015 / Mechanical and Aeronautical Engineering / PhD / Unrestricted

As traffic loads increase and bridges age the need for structural health monitoring is growing. With the digitalization of our society, new non-contact full-field measurement techniques have been developed. These techniques have the potential to be used in monitoring of existing bridges. Today visual inspections are carried out every sixth year. These only give a rough estimate of the structure's health and only provide information about the surface of the structure. In addition to these inspections, traditional sensors like linear variable differential transformers and strain gauges are used to measure parameters such as displacement and strain. For existing bridges in reinforced concrete it is especially important to monitor reinforcement strains, as high strains could be indicative of overloading of the structure or even that a failure is about to occur. The methods available to measure reinforcement strain in existing bridges today are not very effective and have some limitations. The aim of this thesis is thus to evaluate the possibility to predict reinforcement strain based on surface strain measurements obtained by a non-contact full-field optical measurement system. In this study the software ARAMIS was used to measure surface strains, and traditional strain gauges were used to measure reinforcement strain. Strain distribution were evaluated at the initiation of cracks, during sections of cyclic loading and at a load close to the yielding point of the reinforcement. A correlation factor between the strain registered in the software and the strain obtained from the strain gauges was introduced. Based on the results in this study it is not possible to predict exact reinforcement strain based on surface measurements. Digital image correlation does however show potential to be used as a non-contact full-field measurement technique for in-situ measurements. Before this is reality there is still a need for further research in this area.

This thesis investigates the appropriateness of a simplified, open-source digital image correlation (DIC) software for use in quasi-static, structural testing utilizing two-dimensional (2D) DIC measurements. DIC is a non-contact optical measurement technique that uses computer vision to track unique attributes on the surface of an object. For structural testing, traditional instrumentation such as displacement sensors and strain gages are impractical for full field measurements due their limited ability to capture large amounts of data. However, over the past decade, DIC has proven a successful method for full-field kinematics measurements, making it an appealing tool for collecting high densities of accurate data. This thesis specifically studies the accuracy and limitations of the DIC software, MODEM, for various test specimens and loading conditions. This research work is part of an experimental program comprised of three phases. The first stage was conducted by another investigator on aluminum coupons tested in pure tension. These results were used to calibrate parameters (speckle pattern density, lighting, and camera settings) used with the DIC software. The second stage included pure compression tests on concrete cubes and concrete cylinders to compare the difference in results between: (i) surface curvature, (ii) camera distance, (iii) surface treatment, and (iv) speckle pattern color. The final stage involved analysis of a tension test of a concrete prism completed at the University of Auckland in an effort to assess how MODEM could be utilized to accurately detect onset and propagation of concrete cracking. Results showed the most accurate DIC strains were within 5% error when compared to traditional instrumentation for aluminum loaded in tension and within 6% error for concrete loaded in compression/tension. This level of accuracy is comparable to existing open source and commercial DIC software utilizing 2D DIC analysis. Therefore, MODEM can be used to provide accurate 2D DIC strain measurements for small and medium scale structural test specimens when using the following parameters: (i) the surface of the specimen is planar, (ii) the camera is placed accordingly so the maximum amount of zoom can be used, (iii) the surface of a test specimen is free of debris or imperfections, and (iv) a high contrast and evenly distributed speckle pattern is used. Computational analysis of the results showed that known material properties can be used to calibrate, or remove errors from, the DIC results when traditional instrumentation is not available. Additionally, results showed MODEM strain contours can be used for initial detection of cracks in concrete loaded in tension while MODEM tracking performance can be used to characterize the centerline and orientation of cracks. The experimental tests provide critical information on how to set up, run, and analyze DIC results when using MODEM. The full field measurements are of value in providing accurate data for structural testing to develop a better understanding of material response and structural performance, since large-scale tests are typically limited by a sparse number of data points when using traditional instrumentation.