Characteristics Of Main Drive System And Rolling Pressure Prediction Model Of Special Bar And Wire Mill

In view of the phenomenon of large torsional vibration and impact of the main drive system of the rolling mill,excessive fatigue and wear of some transmission parts such as rolls,gears and transmission shafts,and the problem that the rolling pressure can not be accurately displayed in real time,which makes it impossible to judge whether the rolling mill load is safe or not,this paper takes the main drive system of the rolling mill of the special bar as the research object,Using S.Ekelund theoretical formula to calculate the rolling pressure and related process parameters,build a wireless rolling torque measurement and control system,analyze the characteristics of the main drive system of the rolling mill under the existing process conditions,use the finite element simulation software MSC.MARC to analyze the rolling adaptability of the limit specification of the main drive system,improve the RBF neural network,and design a new K-K-RBF neural network rolling pressure prediction model.The main conclusions are as follows:(1)The rolling torque M and rolling pressure P of the main drive system of the special bar and wire mill are uneven,and the working load of the mill is too large.In actual production,impact load is the main reason for the increase of torque amplification factor TAF in the system.The torsional vibration of the upper and lower rolls of the 6V,7H and 8V rolling mills of typical stands is uneven,and the torsional vibration frequency is rich;The torsional vibration frequencies of the upper roll and the lower roll of the 7H rolling mill are obviously inconsistent,which may cause resonance of other transmission parts of the rolling mill.Therefore,it is necessary to strengthen the spot inspection of the reducer of the 6V,7H and 8V rolling mills.(2)MSC.Marc finite element simulation of rolling process shows that the size of finished product rolled in the 8th pass is Φ110mm special bar shall be rolled in the6 th pass Φ126mm.The new rolling process of Φ126mm limit specification special bar has good adaptability,and the rolling torque of each pass mill is close to the design load,realizing the design goal.After the process optimization,the stacking and pulling relationship between the front and rear frames is improved,the temperature field distribution uniformity of the rolled piece section is improved,the distribution uniformity of the equivalent stress and strain of the rolled piece is improved,and the surface defects that may be caused by stress concentration at the finished hole are reduced.The optimized rolling process improves the wear degree of the roll to a certain extent,which can effectively prolong the service life of the roll and improve the production efficiency by reducing the roll change time.(3)The R value of each data set trained,verified and predicted by RBF neural network and K-K-RBF neural network is very high,indicating that the fitting is good and the prediction accuracy is ideal;In particular,the predicted special bar rolling force output by K-K-RBF neural network has a better approximation to the actual special bar rolling force data,the correlation is 0.99,the minimum mean square error is 0.94 KN,and the prediction is reliable.The number of iterations in the whole process is only 46,which effectively reduces the amount of calculation of running the network model.