GOALI: Integrating Micro-scale Physical Chemistry, Fluid-Flow and Process Control for Conceptual Design of a New Aluminum Process

点击翻译按钮获取中文摘要

ABSTRACTPI: B. Erik Ydstie and David Roha Institution: Carnegie Mellon UniversityProposal Number: 0457026Title: GOALI: Integrating Micro-scale, Physical Chemistry, Fluid-Flow and Process Control for Conceptual Design of a New Aluminum ProcessProcesses with integrated flow, chemical reaction and phase transition are used in industry to produce and process a variety of materials. Process simulation can help design and optimize such processes to reduce cost and improve environmental impact parameters. To advance application of this tool however, it is necessary to find new ways to integrate meso-scale fluid flow, micro-scale chemical reactions and phase transformations with macro-scale design and control methods in a stable and consistent manner. The PIs want to develop a computational architecture that combines existing and new software on a distributed cluster computer. Cluster computing allows one to develop process modules and integrate these through network computing in a parallel manner. The specific goal of this research is to develop a computational architecture suitable for modeling, scale-up, design and control of a high-energy chemical reactor for carbothermic production of Aluminum. This reactor has multiple phases, particulates and complex fluid flow. The approach achieves stability and computational scalability by exploiting a symmetry property of physical systems that arises from the second law of thermodynamics. Other examples where this property holds include fluid bed reactors, crystallization, chemical vapor deposition, bio-reactors, slurry reactors, and a large number of processes that rely on particulate processing. This will be part of a multi-disciplinary industry/university program, which aims to develop a coherent design methodology for the carbothermic aluminum process in order to bring it to commercialization within 10 years. Researchers at Carnegie Mellon University and industrial researchers at the ALCOA Technical Center will be involved in the development effort. Intellectual Merit: Multi-scale modeling and distributed computation is of great current interest since they promises to address problems in process design, meteorology, bio-medicine and industry. Advances in this area can give shorter lead times in the development of new processes and products. This research is unique since it combines behavioral constraints imposed by the laws of physics with tools from mathematical systems theory and parallel computing in a novel way. Generic results and theory from this research can be used to design computational tools for integrating new software and legacy code in a distributed multi-scale simulation system.Broader Impacts: Aluminum is an important material for packaging and transportation because of its durability and high strength to weight ratio. This research will help design a new process for making primary aluminum, which reduces the combined capital, energy and operating costs relative to current Hall-Heroult technology by no less than 25%. It will also eliminate the emission of fluoride containing gases. Aluminum in automotive applications reduces transportation costs, relative to steel, due to its low weight.