Optimization of injection-molding process with genetic algorithms

Injection molding is widely used for mass production of polymer products. One important issue is how to determine the gate location(s) and process conditions to produce parts of the best quality. The objective of this paper is to develop an efficient optimization system that can automatically make such determination. The Genetic Algorithm (GA) will be compared with a functional search method, Simulated Annealing (SA) algorithm. The principle of both algorithms will be described and illustrated with examples. Application of these algorithms to determine gate location and optimal process condition in injection molding will be demonstrated with examples. Gate location is determined based on the principle of balanced now paths, while the optimal process condition is computed by minimizing the warpage across the entire part. Results show that the Genetic Algorithm is more efficient computationally than the SA algorithm.