Modelling has become a more and more valuable tool in the design, control and development of steel processing. Empirical regression equations, physically based approachs, artificial neural networks and hybrid models a...Modelling has become a more and more valuable tool in the design, control and development of steel processing. Empirical regression equations, physically based approachs, artificial neural networks and hybrid models are being theied in computer modelling. In all cases, relevant data are necessary, which can be most economically obtained by physical simulation. Physical simulation with a Gleeble simulator has been used in a large number of tasks at the University of Oulu for ten years in cooperotion with the Finnish metals industry. Some examples of these will be described and discussed below, such as the optimization of the recrystallization controlled rolling process, the improvement of the hot strength model for the control of coiling tension and the optimization of continuous strip annealing schedules.Finally,brief remarks will be then on a couple of projects now under way.展开更多
An innovative approach was introduced for the development of a AA6063 recrystallization model.This method incorporated a regression-based technique for the determination of material constants and introduced novel equa...An innovative approach was introduced for the development of a AA6063 recrystallization model.This method incorporated a regression-based technique for the determination of material constants and introduced novel equations for assessing the grain size evolution.Calibration and validation of this methodology involved a combination of experimentally acquired microstructural data from the extrusion of three different AA6063 profiles and results from the simulation using the Qform Extrusion UK finite element code.The outcomes proved the agreement between experimental findings and numerical prediction of the microstructural evolution.The trend of the grain size variation based on different process parameters was accurately simulated,both after dynamic and static recrystallization,with an error of less than 25% in almost the whole sampling computations.展开更多
文摘Modelling has become a more and more valuable tool in the design, control and development of steel processing. Empirical regression equations, physically based approachs, artificial neural networks and hybrid models are being theied in computer modelling. In all cases, relevant data are necessary, which can be most economically obtained by physical simulation. Physical simulation with a Gleeble simulator has been used in a large number of tasks at the University of Oulu for ten years in cooperotion with the Finnish metals industry. Some examples of these will be described and discussed below, such as the optimization of the recrystallization controlled rolling process, the improvement of the hot strength model for the control of coiling tension and the optimization of continuous strip annealing schedules.Finally,brief remarks will be then on a couple of projects now under way.
文摘An innovative approach was introduced for the development of a AA6063 recrystallization model.This method incorporated a regression-based technique for the determination of material constants and introduced novel equations for assessing the grain size evolution.Calibration and validation of this methodology involved a combination of experimentally acquired microstructural data from the extrusion of three different AA6063 profiles and results from the simulation using the Qform Extrusion UK finite element code.The outcomes proved the agreement between experimental findings and numerical prediction of the microstructural evolution.The trend of the grain size variation based on different process parameters was accurately simulated,both after dynamic and static recrystallization,with an error of less than 25% in almost the whole sampling computations.
