The flat rolling process is initiated when the frictional forces draw the strip to be rolled into the roll gap. These forces depend on the coefficient of friction, knowledge of which is essential to understand, describe and analyze the process. Several predictive formulae for the coefficient have been presented in the technical literature. Contradictions are observed, however, when their predictions are compared to each other. The data obtained while cold rolling aluminum and steel strips are used in the analyses. A model of the rolling process – accounting for strain hardening, frictional events and varying speeds - is then used to determine the coefficients of friction. Use of statistical analyses is found to yield more reliable results than the use of the predictive relations.
IntroductionThe success of a mathematical model is judged by the consistency and the accuracy of its ability to predict the variables in the rolling process. Usually these include the roll separating force and the roll torque, determined in carefully controlled experiments. Of the two criteria mentioned above, consistency of the predictive ability is the more important, since sometime accurate predictions are essentially useless. There are many modelling methods published in the technical literature that allows to study the flat rolling process like the empirical models, the one-dimensional models, the upper-bound models and the finite element analysis computing numerous parameters.The empirical model presented by Schey [1] calculates the roll separating force by adjusting the compression of the strip to account for interfacial friction and the relative motion between the roll and the strip. The Hitchcock formula is used for taking into account the roll deformation and it requires an iterative approach. The rolled strip is modelled as rigid -ideally plastic material with isotropic hardening behaviour.The one-dimensional model assumes the presence of plane strain flow, the homogeneous compression to predict the rolling force and torque. For the Bland and Ford's model is widespread to study the cold rolling process [2] The model requires to make several assumptions and simplifications reducing the complexities and allowing the closed form integration of the equation of equilibrium. In the model proposed by Roychoudhuri and Lenard [3] the flattening of the work-roll is taken into account using the two-dimensional theory of elasticity. The strip behaves like an elastic-plastic material which also added a further element of reality to the analysis. In the above mentioned models the coefficient of friction is treated as constant continuously during the rolling process.The upper-bound approach that predict the power required for the steady-state process was also used in the analysis of the flat rolling process [4]. In most treatments the rolls are taken to be rigid and the rolled strips are rigid-ideally plastic, the friction factor is employed constant.Finite-element method provides the opportunity to exam the flat rolling process in details thus numerous applications
Wendelstein 7-X (W7-X) is currently the largest optimized stellarator in operation in the world. Its main objective is to demonstrate long pulse operation and to investigate the suitability of this type of fusion device for a power plant. Maintaining the safety of the first wall is critical to achieving the desired discharge times of approximately 30 min while keeping a steady-state condition. We present a deep learning-based solution to detect the unexpected plasma-wall and plasma-object interactions, so-called hot-spots, in the images of the Event Detection Intelligent Camera (EDICAM) system. These events can pose a serious threat to the safety of the first wall, therefore, to the operation of the device. We show that sufficiently training a neural network with relatively small amounts of data is possible using our approach of mixing the experimental dataset with new images containing so-called synthetic hot-spots generated by us. Diversifying the dataset with synthetic hot-spots increases performance and can make up for the lack of data. The best performing YOLOv5 Small model processes images in 168 ms on average during inference, making it a good candidate for real-time operation. To our knowledge, we are the first ones to be able to detect events in the visible spectrum in stellarators with high accuracy, using neural networks trained on small amounts of data while achieving near-real-time inference times.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.