Numerical algorithms for higher-order accurate discretizations of flows on deforming domains

点击翻译按钮获取中文摘要

Canada is one of the largest producers of hydro and wind energy in the world. Optimally designed marine and wind turbine blades can further increase the production of these sources of renewable energy. More than two hundred floods have occurred in Canada over the past century taking many lives and causing billions of dollars in damage. An accurate prediction of flood plains when rivers overflow may prevent the loss of lives and help protect Canadian homes and businesses. These are just two applications that benefit tremendously from computer simulation tools. The success of these simulation tools, however, depends on the implemented algorithms. In this project we will develop new algorithms for more efficient and more accurate simulation tools.******Recently we introduced a new class of higher-order accurate algorithms for fluid flows on domains with small deformations. This class has the potential of being more accurate and more efficient than many other numerical methods. Many flow simulations, however, require algorithms that can cope with fluid flows on domains with large deformations. Example of such large deformations include: topological changes in which the domain may split into several sub-domains (e.g. coalescence and break-up of bubbles in boiling processes) and flow around rotating components (e.g. in the design of marine or wind turbines). The overarching objective of the proposed research is to develop the methodology needed for simulating fluid flow on domains with large deformations, building upon our higher-order accurate algorithms.******The proposed research benefits Canada by developing algorithms for next generation simulation tools. These tools will enable simulation of a wide variety of flows for applications in industry, environment and society. These applications may include: increasing the production of renewable energy through the design of optimal marine and wind turbine blades; predicting flood plains in the event of rivers overflowing due to extensive rainfall or snow-melt run-off; and predicting blood flow in artificial heart valves. Furthermore, the proposed research will train highly qualified personnel with the critical skills needed to develop and implement accurate numerical methods. This skill-set is essential for software development in Canada as it is widely believed that higher-order numerical methods will soon become the standard in simulation software.