Study On Precise Temperature Control Method Of Hot-rolled Ribbed Bar During Cooling Process

The implementation of the new national standard puts forward new requirements for the performance and metallographic structure of hot-rolled ribbed bar,which brings new opportunities and challenges to the iron and steel industry.It is an urgent issue to be solved for steel enterprises and researchers to find methods of producing hot-rolled ribbed steel bars which meet the requirements of section structure and comprehensive properties,and reducing the production cost to a greater extent.In this paper,the temperature model of hot-rolled ribbed steel bar during cooling process is studied,and the temperature of steel bar during cooling process is controlled precisely according to the characteristics of production.The research results provide sufficient technological optimization support for obtaining qualified microstructure and properties under the condition of moderate reduction of alloy addition.The main contents of this paper are as follows:(1)Based on the basic theory of heat transfer,a mathematical model for temperature prediction of hot-rolled ribbed bar during cooling process is established by using finite difference method.The problem model is simplified to a one-dimensional model,and the explicit and implicit difference formulas are derived by using the finite difference method to calculate the temperature distribution of steel bar under different cooling conditions.(2)The learning method of heat transfer coefficient of hot-rolled ribbed bar during cooling process is established by the data of temperature change in actual production process.Based on the BP neural network architecture,the learning algorithm,the number of hidden layers,input and output parameters are determined.The industrial production data such as pressure,flow rate,water temperature and specifications are used as inputs,and the heat transfer coefficients are used as outputs to train the neural network,so as to realize the adaptive learning of heat transfer coefficients.(3)Aiming at the large amount of production data and the easy interference of the collected data,the data preprocessing algorithm is studied.By comparing the characteristics of data clustering algorithm,the production data are tested,the data are cleaned and screened,the abnormal data are eliminated,the reliable training samples are obtained,the accurate training data are provided for BP neural network,and the prediction accuracy of temperature model is improved.(4)Combined with the equipment arrangement and process regulation in the actual production site,developing an on-line application model for temperature.The temperature field of hot-rolled ribbed bars under different constraint conditions is calculated and the equipment state is set,and the prediction accuracy of the developed temperature field model is verified by instrument measurements,which provides data support for the moderate cooling process of hot rolled ribbed bars after rolling,so as to meet the requirements of the new national standards for the organization and properties of products.In this paper,the temperature of hot-rolled ribbed bar during cooling process is accurately predicted,and the flow rate of the cooler is set and calculated under the condition that the temperature deviation between the surface and the center of the hot-rolled ribbed bar is limited.This method ensures the microstructure of steel bar section and the properties after cooling,which has important theoretical and practical significance for the reduction control of the alloy addition in hot-rolled ribbed steel bar and the homogeneity control of properties.