Thesis (M.S.)--Mississippi State University. Department of Electrical and Computer Engineering. / Title from title screen. Includes bibliographical references.

Thesis (M.S.)--Mississippi State University. Department of Computer Science and Engineering. / Title from title screen. Includes bibliographical references.

Thesis (Ph. D.)--Mississippi State University. Department of Computer Science. / Title from title screen. Includes bibliographical references.

Thesis (Ph. D.)--Mississippi State University. Department of Engineering. / Title from title screen. Includes bibliographical references.

Thesis (M.S.)--Mississippi State University. Department of Computer Science. / Title from title screen. Includes bibliographical references.

Thesis (M.S.)--Mississippi State University. Department of Computer Science and Engineering. / Title from title screen. Includes bibliographical references.

Molecular simulation is an indispensable tool in many different disciplines such as physics, biology, chemical engineering, materials science, drug design, and others. Performing large-scale molecular simulation is of great interest to biologists and chemists, because many important biological and pharmaceutical phenomena can only be observed in very large molecule systems and after sufficiently long time dynamics. On the other hand, molecular simulation methods usually have very steep computational costs, which limits current molecular simulation studies to relatively small systems. The gap between the scale of molecular simulation that existing techniques can handle and the scale of interest has become a major barrier for applying molecular simulation to study real-world problems. In order to study large-scale molecular systems using molecular simulation, it requires developing highly parallel simulation algorithms and constantly adapting the algorithms to rapidly changing high performance computing architectures. However, many existing algorithms and codes for molecular simulation are from more than a decade ago, which were designed for sequential computers or early parallel architectures. They may not scale efficiently and do not fully exploit features of today's hardware. Given the rapid evolution in computer architectures, the time has come to revisit these molecular simulation algorithms and codes. In this thesis, we demonstrate our approach to addressing the computational challenges of large-scale molecular simulation by presenting both the high-performance algorithms and software for two important molecular simulation applications: Hartree-Fock (HF) calculations and hydrodynamics simulations, on highly parallel computer architectures. The algorithms and software presented in this thesis have been used by biologists and chemists to study some problems that were unable to solve using existing codes. The parallel techniques and methods developed in this work can be also applied to other molecular simulation applications.

The purpose of this research was to investigate the role of leadership in creating high performance software development teams. Of specific interest were the challenges faced by the Project Manager without a software engineering background. These challenges included management of a non-visible process, planning projects with significant uncertainty, and working with teams that don’t trust their leadership. Conclusions were drawn from the author’s experience as a software development manager facing these problems and a broad literature review of experts from the software and knowledge worker management fields. The primary conclusion was that, until the next big breakthrough, gains in software development productivity resulting from technology are limited. The only way for a group to distinguish itself as performing at the highest levels is teamwork enabled by good leadership. / text

Simulation of large-scale and complex reservoirs requires fine and detailed gridding, which involves a significant amount of memory and is computationally expensive. Nowadays, clusters of PCs and high-performance computing (HPC) centers are widely available. These systems allow parallel processing, which helps large-scale simulations run faster and more efficient. In this research project, we developed a parallel version of The University of Texas Compositional Simulator (UTCOMP). The parallel UTCOMP is capable of running on both shared and distributed memory parallel computers. This parallelization included all physical features of the original code, such as higher-order finite difference, physical dispersion, and asphaltene precipitation. The parallelization was verified for several case studies using multiple processors. The parallel simulator produces outputs required for visualizing simulation results using the S3graph visualization software. The efficiency of the parallel simulator was assessed in terms of speedup using various numbers of processors. Subsequently, we improved the coding and implementation in the simulator in order to minimize the communications between the processors to improve the parallel efficiency to carry out the simulations. To improve the efficiency of the linear solver in the simulator, we implemented three well-known high-performance parallel solver packages (SAMG, Hypre, and PETSc) in the parallel simulator. Then, the performances of the solver packages were improved in terms of the input parameters for solving large-scale reservoir simulation problems. The developed parallel simulator has expanded the capability of the original code for simulating large-scale reservoir simulation case studies. In other words, with sufficient number of processors, a field-scale simulation with a million grid cells can be performed in few hours. Several case studies are presented to show the performance of the parallel simulator. / text

Polymer nanocomposite foams have attracted tremendous interests due to their multifunctional properties in addition to the inherited lightweight benefit of being foamed materials. Polymer nanocomposite foams using high performance polymer and bio-degradable polymer with carbon nanotubes were fabricated, and the effects of foam density and pore size on properties were characterized. Electrical conductivity modeling of polymer nanocomposite foams was conducted to investigate the effects of density and pore size. High performance polymer Polyetherimide (PEI) and multi-walled carbon nanotube (MWCNT) nanocomposites and their foams were fabricated using solvent-casting and solid-state foaming under different foaming conditions. Addition of MWCNTs has little effect on the storage modulus of the nanocomposites. High glass transition temperature of PEI matrix was maintained in the PEI/MWCNT nanocomposites and foams. Volume electrical conductivities of the nanocomposite foams beyond the percolation threshold were within the range of electro-dissipative materials according to the ANSI/ESD standard, which indicates that these lightweight materials could be suitable for electro-static dissipation applications with high temperature requirements. Biodegradable Polylactic acid (PLA) and MWCNT nanocomposites and their foams were fabricated using melt-blending and solid-state foaming under different foaming conditions. Addition of MWCNTs increased the storage modulus of PLA/MWCNT composites. By foaming, the glass transition temperature increased. Volume electrical conductivities of foams with MWCNT contents beyond the percolation threshold were again within the range of electro-dissipative materials according to the ANSI/ESD standard. The foams with a saturation pressure of 2 MPa and foaming temperature of 100 °C showed a weight reduction of 90% without the sacrifice of electrical conductivity. This result is promising in terms of multi-functionality and material saving. At a given CNT loading expressed as volume percent, the electrical conductivity increased significantly as porosity increased. A Monte-Carlo simulation model was developed to understand and predict the electrical conductivity of polymer/MWCNT nanocomposite foams. Two different foam morphologies were considered, designated as Case 1: volume expansion without nanotube rearrangement, and Case 2: nanotube aggregation in cell walls. Simulation results from unfoamed nanocomposites and the Case 1 model were validated with experimental data. The results were in good agreement with those from PEI/MWCNT nanocomposites and their foams, which had a similar microstructure as modeled in Case 1. Porosity effects on electrical conductivity were investigated for both Case 1 and Case 2 models. There was no porosity effect on electrical conductivity at a given volume percent CNT loading for Case 1. However, for Case 2 the electrical conductivity increased as porosity increased. Pore size effect was investigated using the Case 2 model. As pore size increased, the electrical conductivity also increased. Electrical conductivity prediction of foamed polymer nanocomposites using FEM was performed. The results obtained from FEM were compared with those from the Monte-Carlo simulation method. Feasibility of using FEM to predict the electrical conductivity of foamed polymer nanocomposites was discussed. FEM was able to predict the electrical conductivity of polymer nanocomposite foams represented by the Case 2 model with various porosities. However, it could not capture the pore size effect in the electrical conductivity prediction. The FEM simulation can be utilized to predict the electrical conductivity of Case 2 foams when the percolation threshold is determined by Monte-Carlo simulation to save the computational time. This has only been verified when the pore size is small in the range of a few micrometers. / text