In the quest to advance materials design, this thesis integrates Machine Learning (ML) techniques with Density Functional Theory (DFT) data. A novel representation called splashdown is formulated to capture long-range interactions, an aspect often neglected by material representations. A project known as ORGANIZER leads to the creation of a pivotal database, culminating in the discovery of a new organic solid-state lasing molecule that doubled the state-of-the-art emission gain cross-section. Concurrently, a monte-carlo based optimizer, aMC, is tested, demonstrating superior performance to gradient-based methods without the need for expensive gradient computation. Enhanced Graph Neural Networks (GNN)s predict High Entropy Alloy (HEA) catalysts for oxygen reduction reaction, halving necessary DFT computations and unveiling a new HEA catalyst with a 0.27V overpotential. The splashdown representation compares to state-of-the-art ones like MBTR and SOAP in predicting long-range interactions. Collectively, these efforts highlight the transformative potential of ML and some adjacent fields in materials science.

Since the Conform process was patented 30 years ago, there have only been approximately 200 machines sold worldwide. Given that Conform competes economically with conventional extrusion and is also reported to be a more energy efficient process, it is surprising that the use of Conform is not more widespread in today's increasingly environmentally conscious and high-production focussed world. One explanation for this is likely to be due to the fact that there is still limited knowledge of the thermo-mechanical behaviour of the workpiece during extrusion. Furthermore, for the aluminium industry, there are still issues remaining regarding the production of flash and the quality of the extrudate in terms of mechanical properties. This study provides the reader with the findings of the research and experimental work undertaken by the author, his co-workers and fellow specialists, in the field of aluminium extrusion including Conform. The experimental work includes both laboratory experiments performed with a direct extrusion press and an experimental machine set up to replicate the Conform process. The experimental work is also simulated using finite element modelling techniques. The results from these analyses are then validated by comparing industrial and experimental data. The finite element analyses are enhanced by using parallel processing technology and user sub-routines. The author proposes new models to allow for the study of the different sub-processes in Conform. These include the coining of the feedstock, formation of the upset zone, extrusion of the flash, the filling-up of the expansion chamber / feeder plate and the extrusion of the extrudate. The author also investigates methods which predict microstructure and surface cracks in the extrudate. The author suggests innovative techniques to improve the efficiency of finite element analysis in metal forming. Finally the author recommends procedures for the study of structural integrity and the optimisation of the tooling used in Conform.

published_or_final_version / abstract / Mechanical Engineering / Master / Master of Philosophy

No description available.

SC-15 epoxy is used in many industrial applications and it is well known that the mechanical and viscoelastic properties of epoxy can be signicantly enhanced when reinforced with nanofillers. In this work, SC-15 epoxy is reinforced by loading with magnetically-active nanofillers and cured in a modest magnetic field. Because of the signicant magnetic response of the nanofillers, this is a low cost and relatively easy technique for imposing a strong magnetic anisotropy to the system without the need of a superconducting magnet. It is also found that this method is an effective way of enhancing the mechanical properties of epoxy. Three systems were prepared and studied. The first is a dilute system of various concentrations of Fe2O3 nanoparticles in SC-15 epoxy. The second system is a combination of Fe2O3 nanoparticles and chemically-functionalized single-walled carbon nanotubes (SWCNT(COOH)s) in SC-15 epoxy. The third is a dilute system of SWCNT(COOH)s decorated with Fe3O4 particles t hrough a sonochemical oxidation process in SC-15 epoxy. Samples have an initial cure of 6 hrs in a magnetic led of 10 kOe followed by an additional 24 hours of post curing at room temperature. These are compared to the control samples that do not have initial field curing. Tensile and compressive stress-strain analysis of the prepared systems shows that mechanical properties such as tensile strength, tensile modulus and compressive strength are enhanced with the inclusion of these nanofillers. It is also found that there is an anisotropic enhancement of these properties with respect to the imposed curing field. An interesting phenomenon is observed with the increase in modulus of toughness and fracture strain with nanotube inclusion. / These parameters are drastically enhanced after curing the systems in a magnetic field. While there is a modest shift in glass transition temperature during viscoelastic analysis, the thermal stability of the created systems is not compromised. Results of these mechanical enhancements will be compared with other nanoloading techniques from literature. / by Olga Malkina. / Thesis (Ph.D.)--Florida Atlantic University, 2011. / Includes bibliography. / Electronic reproduction. Boca Raton, Fla., 2011. Mode of access: World Wide Web.

published_or_final_version / abstract / toc / Orthopaedics and Traumatology / Master / Master of Philosophy

No description available.

In this article-based thesis, we cover applications of deep learning to different problems in condensed matter physics, where the goal is to either accelerate the computation or design of a nanoscale material. We first motivate and introduce how machine learning methods can be used to accelerate traditional condensed matter physics calculations. In addition, we discuss what designing a material means, and how it has been previously done. We then consider the fundamentals of electronic structure and conventional calculations which include density functional theory (DFT), density functional perturbation theory (DFPT), quantum Monte Carlo (QMC), and electron transport with tight binding. In addition, we cover the basics of deep learning. Afterwards, we discuss 6 articles. The first 5 articles are dedicated to accelerating the computation of nanoscale materials. In Article 1, we use convolutional neural networks to predict energies for diatomic molecules modelled with a Lennard-Jones potential and density functional theory energies of hexagonal lattices with and without defects. In Article 2, we use extensive deep neural networks to represent density functional theory energy functionals for electron gases by using the electron density as input and bypass the Kohn-Sham equations by using the external potential as input. In addition, we use deep convolutional inverse graphics networks to map the external potential directly to the electron density. In Article 3, we use voxel deep neural networks (VDNNs) to map electron densities to kinetic energy densities and functional derivatives of the kinetic energies for graphene lattices. We also use VDNNs to calculate an electron density from a direct minimization calculation and introduce a Monte Carlo based solver that avoids taking a functional derivative altogether. In Article 4, we use a deep learning framework to predict the polarization, dielectric function, Born-effective charges, longitudinal optical transverse optical splitting, Raman tensors, and Raman spectra for 2 crystalline systems. In Article 5, we use VDNNs to map DFT electron densities to QMC energy densities for graphene systems, and compute the energy barrier associated with forming a Stone-Wales defect. In Article 6, we design a graphene-based quantum transducer that has the ability to physically split valley currents by controlling the pn-doping of the lattice sites. The design is guided by an neural network that operates on a pristine lattice and outputs a lattice with pn-doping such that valley currents are optimally split. Lastly, we summarize the thesis and outline future work.

In this investigation, polymer precursor of syntactic foam has been reinforced with SiC nanoparticles to enhance mechanical and fracture properties. Derakane 8084 vinyl ester resin was first dispersed with 1.0 wt% of SiC particles using a sonic cavitation technique. In the next step, 30.0 wt% of microspheres (3M hollow glass borosilicate, S-series) were mechanically mixed with the nanophased vinyl ester resin, and cast into rectangular molds. A small amount of styrene was used as dilutant to facilitate mixing of microspheres. The size of microspheres and SiC nanoparticles were 20-30 um and 30-50 nm, respectively. Tension, compression, and flexure tests were conducted following ASTM standards and a consistent improvement in strength and modulus within 20-35% range was observed. Fracture toughness parameters such as KIC and GIC were also determined using ASTM E-399. An improvement of about 11-15% was observed. Samples were also subjected to various environmental conditions and degradation in material properties is reported. / by Debdutta Das. / Thesis (M.S.C.S.)--Florida Atlantic University, 2009. / Includes bibliography. / Electronic reproduction. Boca Raton, Fla., 2009. Mode of access: World Wide Web.

published_or_final_version / abstract / toc / Mechanical Engineering / Doctoral / Doctor of Philosophy