Finite Element Analysis and Prediction of the Cross-Section Morphology of Grid Strips

To investigate the relationships between the quality of the section morphology of the zirconium alloy strips and the blanking parameters as well as the design of the blanking mould, 64 processing parameter combinations are generated by using Latin Hypercube Sampling and the finite element models with different boundary conditions have been constructed to acquire the corresponding section qualities. The calculation results of the finite element model are verified by the experiment data. Furthermore, the random forest algorithm is employed to build the prediction model between the processing parameters and the lengths of the section bright strip. The sensitive analysis of the processing parameters influenced on the section quality have been conducted through the prediction model. The results indicate that the random forest prediction model have a satisfied accuracy, which could effectively forecast the length of the bright band in the section, and then provides guidance on mold design as well as process parameter selections. The blanking clearance included in the design factor has a decisive influence on the length of the bright strip; The friction coefficient has the weakest effect on the results. The effect of die fillet radius and blanking speed on the length of bright band is not significant.