Division of Information Technology, Engineering and the Environment
School of Advanced Manufacturing and Mechanical Engineering
School of Advanced Manufacturing and Mechanical Engineering
Thesis Abstract
Hot steel rolling is amongst the most important metalworking processes because of huge amount of consumed resources, high efficiency, immense environmental impact, as well as the significance and enormous quantity of long products. According to the final profile of products, hot steel rolling can be classified into flat rolling and shape rolling. As the kernel of shape rolling, roll pass design (RPD) has great impact on products quality, productivity and working life of rolling systems, as well as energy consumption of the rolling operations. Thus, the optimal design of roll pass has profound influence on the optimization of the overall manufacturing system.
The acceleration of industrialization has seen a rapid increase of demand for steel products, placing heavy burdens on energy supply and the environment. To satisfy the ever-increasing demand in the market and competition in the steel industry, manufacturers are facing the challenges from technological innovation, product quality, process flexibility and productivity, as well as energy and other resources consumption. In order to achieve these on the industrial scale, numerous combinations of system parameters must be analyzed, and RPD in particular must be optimized. The intricate relationships between parameters greatly increase the complexity of the design models, and place heavy computational burdens on designers as well as computer-aided systems. It is widely accepted that even for a simple cross section rolling, relationships between parameters are remaining highly uncertain. Thus, it is difficult to get an efficient solution without losing its accuracy through any single-disciplinary knowledge and single model.
This study aims to present a parameterized generalized strategy, which is based on a hybrid model and cross-disciplinary knowledge, for roll pass optimal design. The disciplines which must be included range from engineering design, mechanical metallurgy and plastic forming of steels to operations research, stochastic modeling and maintenance management. Design vectors anticipated in this model are based on both geometric parameters of roll passes and physical parameters of the work piece and rolls. In order to overcome the limitations of the isolated modeling and analysis, a combination of strategies, such as statistical approaches, evolution algorithms, and finite element methods are used for analysis and optimization. Lab-scale rolling is used for methodology verification. Challenges lie in how to choose different types of parameters to describe the design space and how to integrate different modeling and analysis strategies. Solutions will be applied to initial passes in an industrial rolling mill.
