Purpose: Punching process currently plays an important role in industrial mass production. The current study focuses on increasing accuracy, performance and extending service life of punches and dies. Optimizing the design of the punch and die has been a common topic for scholars. Design/methodology/approach: The input parameters (punching times, clearance) and output results (wear) were used to construct a training database. The abductive network formulation established a relationship between input parameters and output results. By using the abductive modeling technique, the complicated and uncertain relationships between the input and output variables can be formulated into a useful mathematical model. Once the abductive network model was constructed, the relationships between input and output parameters variables became available. A simulated annealing algorithm (SA) with a performance index was established to optimize this process and find the best result as compared with the actual experiment values. Findings: This study aims to establish the relationship between punching times, clearance and wear of micro punches using the abductive network, and to find relational model involving input parameters and output result of punching die in composite blanking processes. This model can be used to estimate wear between punch and die for industrial applications. Research limitations/implications: Setting up the relational expression of punching times, clearance and wear requires a well-established database covering sufficient relevant parameters and data. In training the database, it is helpful to establish a good relational model among punching parameters. Incorrect data will cause abnormal wear. As a result, the mathematics model is difficult to converge and the neural network will inaccurately predict wear. In addition, the punching die may be changed prematurely, which increases production costs. Delay in replacing a worn punching die can result in poor quality of products. Originality/value: As electronic production becomes increasingly smaller, the opportunity to use micro punches and dies will expand accordingly. This study established the relational expression of input and output, which can be used to correctly estimate any wear condition. This result is based on an abductive network and SA method.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.