Application of the strain localization CAFE model to investigate extrusion with various die shapes

Multi scale CAFE model for the prediction of initiation and propagation of the micro shear bands and shear bands in metallic materials subjected to plastic deformation is presented. The CAFE approach is the combination of the Cellular Automata (CA) and the Finite Element (FE) methods. The application of the developed CAFE model to analyze material flow during extrusion is the objective of the present work. The proposed CAFE approach is applied in this work to simulation of the extrusion with flat face and convex dies and to investigate differences in the material flow. The initial FE meshes with the set of the CA point are generated for the numerical tests and the results of the metal flow predicted by the CAFE method are presented in the paper.

Modeling of advanced high strength steels with the realistic microstructure–strength relationships

The objective of the work is to consider the first-order effects of the realistic microstructure morphology in the macroscale modeling of the multiphase Advanced High Strength Steels (AHSS). Instead of using constitutive equations at macroscale, the strength–microstructure relationship is studied in the forms of micromechanical and multiscale models that do not make considerable simplifications with regard to the microscale geometry and topology. The trade-off between the higher computational time and the higher accuracy has been offset with a stochastic approach in the construction of the microscale models. The multiphase composite effects of AHSS microstructure is considered in realistic microstructural models that are stochastically built from AHSS micrographs. Computational homogenization routines are used to couple micro and macroscale and resultant stress–strain relations are compared for models built with the simplified and idealized geometries of the microstructure. The results from this study show that using a realistic representation of the microstructure, either for DP or TRIP steel, could improve the accuracy of the predicted stress and strain distribution. The resultant globally averaged effective stress and strain fields from realistic microstructure model were able to accurately capture the onset of the plastic instability in the DP steel. It is shown that the macroscale mechanical behavior is directly affected by the level of complexities in the microscale models. Therefore, greater accuracy could be achieved if these stochastic realistic microstructures are used at the microscale models.

Stereo image matching using wavelet scale-space representation

A multi-resolution technique for matching a stereo pair of images based on translation invariant discrete multi-wavelet transform is presented. The technique uses the well known coarse to fine strategy, involving the calculation of matching points at the coarsest level with consequent refinement up to the finest level. Vector coefficients of the wavelet transform modulus are used as matching features, where modulus maxima defines the shift invariant high-level features (multiscale edges) with phase pointing to the normal of the feature surface. The technique addresses the estimation of optimal corresponding points and the corresponding 2D disparity maps. Illuminative variation that can exist between the perspective views of the same scene is controlled using scale normalization at each decomposition level by dividing the details space coefficients with approximation space and then using normalized correlation. The problem of ambiguity, explicitly, and occlusion, implicitly, is addressed by using a geometric topological refinement procedure and symbolic tagging.

Advanced modelling of forming high strength steels

The thesis presents a novel approach in the multiscale modelling of Advanced High Strength Steels for prediction of the microstructural effects in forming processes. The results are compared with that of experiments and finite element method. The method is proved to be suitable for complexities in the multiphase AHSS.

A new solution method for homogenization of effective properties of electromagnetic honeycombs

In this paper, an analytical model and its new numerical solution using the homogenization method are developed to determine the effective electromagnetic characteristics of honeycombs. Based on the proposed solution method, the electromagnetic properties are obtained by employing the multi-scale homogenization theory and periodical electric (magnetic) potential boundary conditions. Further, the effect of geometry of honeycomb's unit cell on effective electromagnetic properties is investigated with the use of the proposed method. The numerical results are compared with analytic results using the Smith-Scarpa's semi-empirical formula.

Synthesis and characterization of silver nanoparticles and titanium oxide nanofibers : toward multifibrous nanocomposites

A new method was investigated to produce new multiscale fibrous nanocomposites comprised of titanium oxide (TiO₂) nanofibers and silver (Ag) nanoparticles (NPs). The process involved electrospinning TiO₂ precursor solution containing colloidal solution of Ag NPs, and organic solvent (dimethyl-n′n-formamide) to fabricate a porous, nonwoven, free-standing nanofiber mesh. Postprocess heating of the electrospun nanofibers entailed calcination in air environment at 500°C for 3 h. Microemulsion processing was used to generate NPs of Ag in a monodispersed distribution throughout the colloidal solution. X-ray diffraction data were consistent with the anatase phase of TiO2, while transmission electron microscopy and hydrogen desorption measurements revealed a very porous microstructure. It was demonstrated that NP colloidal stability is solvent dependent. It is anticipated that incorporation of metal particles in nanofibers will lead to enhanced photocurrent generation, when used in functional devices.

Strain partitioning in dual-phase steels containing tempered martensite

Tempering has been used as a method to develop a range of dual phase steels with the same martensite morphology and volume fraction, but containing phases with different relative strengths. These steels were used to examine the strain partitioning between the two constituent phases experimentally through mechanical testing and numerically through finite element modelling. It was found that increasing the differential in strength between the two phases not only produces regions of high strain, but also regions of low strain. On average, a larger difference in strength between the phases increased the strain carried by the softer phase. There was no discernible preferential strain localisation to the ferrite/martensite interface, with the regions of strain localisation being determined by the morphology of the microstructure. A direct correlation between the average strain in the ferrite, and the measured ductility has been found. © 2014 Elsevier B.V.

The development of ultrafine-grained hot rolling products using advanced thermomechanical processing

The aim of the work is development of industry guidance concerning production of ultrafine-grained (UFG) High Strength Low Alloy (HSLA) steels using strain-induced dynamic phase transformations during advanced thermomechanical processing. In the first part of the work, the effect of processing parameters on the grain refinement was studied. Based on the obtained results, a multiscale computer model was developed in the second part of the work that was subsequently used to predict the mechanical response of studied structures. As an overall outcome, a process window was established for the production of UFG steels that can be adopted in existing hot rolling mills. © 2014 Elsevier B.V.

Microstructure modeling and prediction of the mechanical properties of advanced high strength steels

Dual phase (DP) steels were modeled using 2D and 3D representative volume elements (RVE). Both the 2D and 3D models were generated using the Monte-Carlo-Potts method to represent the realistic microstructural details. In the 2D model, a balance between computational efficiency and required accuracy in truly representing heterogeneous microstructure was achieved. In the 3D model, a stochastic template was used to generate a model with high spatial fidelity. The 2D model proved to be efficient for characterization of the mechanical properties of a DP steel where the effect of phase distribution, morphology and strain partitioning was studied. In contrast, the current 3D modeling technique was inefficient and impractical due to significant time cost. It is shown that the newly proposed 2D model generation technique is versatile and sufficiently accurate to capture mechanical properties of steels with heterogeneous microstructure.

DeepCare: a deep dynamic memory model for predictive medicine

Personalized predictive medicine necessitates the modeling of patient illness and care processes, which inherently have long-term temporal dependencies. Healthcare observations, recorded in electronic medical records, are episodic and irregular in time. We introduce DeepCare, an end-to-end deep dynamic neural network that reads medical records, stores previous illness history, infers current illness states and predicts future medical outcomes. At the data level, DeepCare represents care episodes as vectors in space, models patient health state trajectories through explicit memory of historical records. Built on Long Short-Term Memory (LSTM), DeepCare introduces time parameterizations to handle irregular timed events by moderating the forgetting and consolidation of memory cells. DeepCare also incorporates medical interventions that change the course of illness and shape future medical risk. Moving up to the health state level, historical and present health states are then aggregated through multiscale temporal pooling, before passing through a neural network that estimates future outcomes. We demonstrate the efficacy of DeepCare for disease progression modeling, intervention recommendation, and future risk prediction. On two important cohorts with heavy social and economic burden -- diabetes and mental health -- the results show improved modeling and risk prediction accuracy.

Associations between step duration variability and inertial measurement unit derived gait characteristics

Inertial measurement units (IMU) provide a convenient tool for gait stability assessment. However, it is unclear how various gait characteristics relate to each other and whether gait characteristics can be obtained from resultant acceleration. Therefore, step duration variability was measured in treadmill walking from 39 young ambulant volunteers (age 24.2 [± 2.5] y; height 1.79 [± 0.09] m; mass 71.6 [± 12.0] kg) using motion capture. Accelerations and gyrations were simultaneously recorded with an IMU. Harmonic ratio, maximum Lyapunov exponents, and multiscale sample entropy (MSE) were calculated. Step duration variability was positively associated with MSE with coarseness levels = 3-6 (r = -.33 to -.42, P ≤ .045). Harmonic ratio and MSE with all coarseness levels were negatively associated (r = -.45 to -.57, P ≤ .004). The MSE with coarseness level = 2 was negatively associated with short-term maximum Lyapunov exponents (r = -.32, P = .047). The agreement between resultant and vertical acceleration derived gait characteristics was excellent (ICC = 0.97-0.99). In conclusion, MSE with varying coarseness levels was associated with the other gait characteristics evaluated in the study. Resultant and vertical acceleration derived results had excellent agreement, which suggests that resultant acceleration is a viable alternative to considering the acceleration dimensions independently.