Roll Wear Prediction Model For Finishing Trains Of Hot Strip Mill Based On Mechanism And Data

The work rolls of finishing mills on the modern strip hot rolling production lines have the characteristics of high rolling speed and large load bearing.Although rolls wear of the finishing mills is reduced due to the application of the new materials and new processes for rolls,uneven wear of the rolls is still severe,which has a great influence on the quality of thin strips especially.So frequent shutdowns for replacement of the worn rolls are required,which severely limits production efficiency.The strategy named inline roll change can realize roll change without stopping the machine,witch can help to improve production efficiency of hot strip rolling line.The prerequisite for the implementation of this strategy is to determine the roll change time point,and the key step is to accurately predict the amount of roll wear,and then the impact of roll wear on the strip quality should also be clear.The prerequisite for the implementation of this strategy is to accurately predict the amount of the rolls wear.The existing mechanism model can predict the roll wear trend,but it is weak in predicting accuracy.The prediction model based on data does not consider the mechanism of roll wear,so it is weak in predicting stability.In this paper,a roll wear prediction model based on mechanism and data is established,which uses mechanism prediction model as a framework and introduces field production data.It is verified that the prediction model has good performance in roll wear prediction accuracy and stability.At the same time,the roll wear degree corresponding to the roll change time node is obtained.The main contents of this paper are as follows:Firstly,the roll wear mechanism prediction model was derived,and the lack of its prediction accuracy is pointed out through simulation analysis.Taking the roll wear mechanism prediction model as the framework,considering the influence of rolling temperature and rolling speed,and introducing various influencing parameters that should be solved with field production data,the prediction model of roll wear based on mechanism and data was established.Secondly,based on the data of 2250 mm production line,the key parameters of the model based on mechanism and data were determined separately,including rolling force distribution,contact force distribution between work roll and back roll,work roll radius distribution,relative sliding distance between roll and plates and model influence parameters.After that,the accuracy and stability of the prediction model were verified using different rolling schedules.Thirdly,based on the data of 2250 mm production line,the influence of work roll wear in hot strip finishing mill was studied to obtain the influence laws of work roll under uniform and uneven wear states on strip thickness,strip crown and strip shape.Then the relevant problems of determining the time node of inline roll change are analyzed,and the wear degree corresponding to the time node of the work roll change of the F7 finishing stand on the hot strip production line is determined.Finally,the roll wear prediction software was developed based on the C# language to achieve the prediction of wear amount and wear degree of the work rolls in hot strip rolling finishing mills.